Flow preferences of benthic macroinvertebrates in three

Flow preferences of benthic macroinvertebrates in three

Scottish Rivers

by

Matthew Thomas OHare

This thesis is submitted for the degree of Doctor of Philosophy, Division of Environmental & Evolutionary Biology, Insitute of Biomedical & Life

Sciences and the Department of Civil Engineering, University of Glasgow,September 1999.

© Matthew T. OHare, September 1999.

ProQuest Number: 13818646

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a com p le te manuscript and there are missing pages, these will be noted. Also, if material had to be removed,

a note will indicate the deletion.

uestProQuest 13818646

Published by ProQuest LLC(2018). Copyright of the Dissertation is held by the Author.

All rights reserved.This work is protected against unauthorized copying under Title 17, United States C ode

Microform Edition © ProQuest LLC.

ProQuest LLC.789 East Eisenhower Parkway

P.O. Box 1346 Ann Arbor, Ml 48106- 1346

G IASG OWUNIVERSITYLIBRARY

This thesis is dedicated to my parents.

Abstract

Scottish freshwaters have been described as a national resource of international significance. The high quality of Scotland’s lotic systems is integral to the formation of this view. The research presented here aims to provide an insight into the interaction between benthic invertebrates and their hydraulic habitat within some of Scotland’s lotic systems. A further aim of this project is that this information presented here will aid the design of river rehabilitation and management schemes thereby helping maintain the integrity of the opening statement.

There is a large amount o f literature existing which addresses the interactions between benthic invertebrates and flow parameters; substrate type, vegetation, velocity, depth and near bed stresses. However significant gaps remain in our understanding, particularly at the level of individual taxa preferences. Furthermore, little work has been done in Scotland. To address these gap in the data the distribution of macroinvertebrates in relation to flow parameters were assessed for three rivers representative of highland (River Etive), central belt (Blane Water) and borders rivers (Duneaton Water).

The importance of deep and shallow reaches as habitat units for benthic invertebrates was analysed and the methods for categorising reaches into riffles, runs and pools assessed. The analysis showed that at the sites examined differences between invertebrate community in deep and shallow reaches were minimal and limited to the preferences of a number of key species. Categorising reaches into riffles, runs and pools on purely visual grounds was insufficient and some measures of velocity and depth are required if the work is to be used for between site comparisons.

Benthic invertebrates did show preferences for flow parameters. At the physical scale examined (Surber sample) community structure was influenced in a limited manner by flow parameters; velocity and depth wre the most important. A gradient from erosional to depositional conditions was observed at two of the sites.

Limitations of Instream Flow Incremental Methodology (IFIM) as applied to benthic invertebrate habitat identification were identified. Estimates of near bed flow parameters based on point measurements of velocity profiles to samples collected at the scale of Surber samples do not explain any additional variation in the distribution of benthic invertebrates. Analysis of individual flow preferences o f macroinvertebrates suggest that to identify flow preference curves, an aim of IFIM, finer scale habitat measurements are needed.

Laboratory experiments were carried out to identify the upper velocity tolerances of some benthic invertebrates; Tipulidae and Gammarus pulex. The results show that individuals were flexible in their responses to high velocities. What constituted ‘high’ velocity was taxa specific.

Benthic invertebrate community structure was investigated in areas of the Blane Water vegetated with Callitriche instagnalis. Submerged vegetated patches supported a

greater abundance of invertebrates than bare substrate. The hydraulic habitat of the macrophyte stands was more diverse than that of bare substrate with higher velocities occurring on the outside of the macrophyte stands than on the bare substrate. Simuliidae dominated the outside of the stands, the area exposed to the highest velocities. The invertebrate community on the outside o f the plant stands was less equitable than that found at the root-substrate interface. It is suggested that macrophytes can be used as a tool in the rehabilitation of hydraulic habitat for benthic invertebrates in Scottish rivers.

The importance of these results are discussed in the context of river rehabilitation and our ecological understanding of benthic invertebrate community structure.

CONTENTS

ABSTRACT..................................................................................................................................................................... i

CONTENTS................................................................................................................................................................... iii

DECLARATION...........................................................................................................................................................vi

ACKNOWLEDGEMENTS.......................................................................................................................................vii

CHAPTER 1 GENERAL INTRODUCTION........................................................................................................... 1

1.1 T h e b io l o g y o f fl o w in g -w a t e r b e n t h ic m a c r o -in v e r t e b r a t e s ................................................................1

1.2 L o t ic s y s t e m s : P h y sic a l s t r u c t u r e ........................................................................................................................ 71.2.1 Life, light, temperature & water chemistry............................................................................................ 81.2.2 Drainage networks and Channel Structure.......................................................................................... 101.2.3 Scale o f physical processes, implications fo r ecology........................................................................12

1.3 L o t ic s y s t e m s : e c o l o g ic a l in t e r p r e t a t io n .......................................................................................................13

1.4 Spec ific s o f St u d y ........................................................................................................................................................... 17

1.5 T h e sis o u t l in e ....................................................................................................................................................................23

CHAPTER 2: HYDRAULIC AND INVERTEBRATE SURVEYS OF REACHES IN THE BLANE WATER, RIVER ETIVE AND DUNEATON WATER ................................................................................................. 25

2.1 In t r o d u c t io n ..................................................................................................................................................................... 25

2.2 M e t h o d s ................................................................................................................................................................................ 302.2.1 Site choice ..................................................................................................................................................302.2.2 Description o f s ite s ...................................................................................................................................312.2.3 Field measurements..................................................................................................................................332.2.4 Estimation o f hydraulic parameters from velocity profiles............................................................... 342.2.5 Statistical analysis.................................................................................................................................... 36

2.3 R e s u l t s ...................................................................................................................................................................................382.3.1 General Conditions...................................................................................................................................382.3.2 Comparison o f flow conditions at different sites ................................................................................ 402.3.3 Substrate available in the reaches......................................................................................................... 442.3.4 Distribution o f Turbulent and Laminar flow within reaches.............................................................482.3.5 Species lists.................................................................................................................................................502.3.6 Reach preferences o f taxa ....................................................................................................................... 53

2 .4 D is c u s s io n ............................................................................................................................................................................ 542.4.1 Physical character o f the Sections......................................................................................................... 552.4.2 Biological character o f the Sections......................................................................................................56

2.5 C o n c l u s io n s ........................................................................................................................................................................ 602.5.1 Physical study ............................................................................................................................................ 602.5.2 Biological study .........................................................................................................................................61

CHAPTER 3: BENTHIC INVERTEBRATE COMMUNITY ORDINATION......................................................63

3.1 In t r o d u c t io n .................................................................................................................................................................. 633.1.1 Outline...................................................................................................................................................... 63

3.1.2 Physical patchiness.................................................................................................................................. 633.1.3 Spatial aggregations................................................................................................................................ 643.1.4 Assessment o f the ecohydrological health o f rivers............................................................................64

3.2 Methods............................................................................................................................................................ 673.2.1 Data Collection and Dataset Structures.............................................................................................. 673.2.2 Analysis procedure....................................................................................................................................68

3.3 Results...............................................................................................................................................................723.3.1 Suitability o f data sets .............................................................................................................................. 723.3.2 River E tive .................................................................................................................................................. 733.3.3 Duneaton Water.........................................................................................................................................763.3.4 Blane Water................................................................................................................................................81

3.4 Discussion......................................................................................................................................................... 843.5 Conclusions..................................................................................................................................................... 88

CHAPTER 4: INDIVIDUAL RESPONSES TO HYDRAULIC PARAMETERS........................................... 90

4.1 Introduction................................................................................................................................................... 90

4.2 Methods............................................................................................................................................................ 934.2.1 Data source.................................................................................................................................................934.2.2 Statistical analysis....................................................................................................................................93

4.3 Results...............................................................................................................................................................984.3.1 Curve fitting ................................................................................................................................................984.3.2 Relationships o f totalled maximum abundance to the frequency o f variable intervals................984.3.3 Relationships o f maximum sample abundance to the frequency o f variable intervals............. 1034.3.4 Velocity.................................................................................................................................................... I l l4.3.5 Depth ........................................................................................................................................................ 1204.3.6 Substrate....................................................................................................................................................129

4.4 D iscussion....................................................................................................................................................... 130

4.5 Conclusions....................................................................................................................................................132

CHAPTER 5: FLUME EXPERIMENTS: ENTRAINMENT VELOCITIES OF BENTHIC INVERTEBRATES....................................................................................................................................................133

5.1 Introduction................................................................................................................................................. 133

5.2 Materials....................................................................................................................................................... 1385.2.1 Flumes...................................................................................................................................................... 1385.2.2 Flow Structure in the flum e ................................................................................................................. 1415.2.3 Experimental animals........................................................................................................................... 143

5.3 Methods...........................................................................................................................................................1445.3.1 Weighing and morphometric measurements..................................................................................... 1445.3.2 Experimental design and statistical analysis................................................................................... 146

5.4 Results............................................................................................................................................................. 1505.4.1 Head width to frontal area ra tios ...................................................................................................... 1505.4.2 Phototaxis................................................................................................................................................ 1505.4.3 Rheotaxis and swimming...................................................................................................................... 1515.4.4 Entrainment logistic curves................................................................................................................. 154

iv

5.4.5 Effect o f weight on entrainment ofTipulidae ................................................................................ 157

5.5 D is c u s s io n ....................................................................................................................................................... 158

5.6 C o n c lu s i o n s ....................................................................................................................................................162

CHAPTER 6: INVERTEBRATE HYDRAULIC MICROHABITAT AND COMMUNITY STRUCTURE IN CALLITRICHE STAGNALIS SCOP. PATCHES.................................................................163

6.1 I n t r o d u c t i o n ..................................................................................................................................................1636.2 M e t h o d s ........................................................................................................................................................... 164

6.2.1 Source o f p lan ts ..................................................................................................................................... 1646.2.2 Measurements...........................................................................................................................................1656.2.3 Data manipulation................................................................................................................................. 1666.2.4 Statistical analysis...................................................................................................................................166

6.3 R e s u l t s ............................................................................................................................................................. 1686.3.1 Comparisons between plant and Surber sam ples............................................................................ 1686.3.2 Comparisons between plant sections................................................................................................. 1686.3.3 Community structure............................................................................................................................. 1736.3.4 Prevalent flow conditions.......................................................................................................................176

6.4 D is c u s s io n ........................................................................................................................................................1776.5 C o n c lu s i o n s .................................................................................................................................................... 179

CHAPTER 7: GENERAL DISCUSSION.............................................................................................................. 181

7.1 R e v ie w o f r e s u l t s ...........................................................................................................................................................181

7.2 F u t u r e W o r k .................................................................................................................................................................... 182

7.3 M a n a g e m e n t r e c o m m e n d a t io n s ............................................................................................................................ 184

REFERENCES............................................................................................................................................................ 189

APPENDIX I: DEFINITIONS OF FLOW UNITS AND PARAMETERS......................................................205

APPENDIX II: EXAMPLE OF LOG PLOTTING OF VELOCITY PROFILES WITH NOTES ON DISTRIBUTION PATTERNS OF SIGNIFICANT PROFILES........................................................................ 206

APPENDIX III: VELOCITY PROFILES. RAW DATA AND REGRESSION ANALYSIS .................... 212

APPENDIX IV: TWINSPAN ORDERED SPECIES BY SAMPLES TABLES........................................... 228

APPENDIX V: INDIVIDUAL RESPONSES OF BENTHIC INVERTEBRATES TO VELOCITY.DEPTH AND SUBSTRATE.................................................................................................................................... 233

Declaration

I declare that the work described in this thesis has been carried out by myself unless otherwise acknowledged. It is entirely of my own composition and has not, in whole or part, been submitted for any other degree.

Matthew OHare September 1999

Acknowledgements

First of all I would like to gratefully thank my supervisors, Dr Kevin

Murphy, Dr. Alan Taylor and Prof. D. Alan Ervine, for having the

patience to let the study go its own way whilst offering excellent

supervision and encouragement throughout.

I lived at the University Field Station, Rowardennan for the majority of the

project, so I would like to gratefully thank the Field Station folk for their

kindness, encouragement, help and good humour; Dr. Colin Adams, Dr.

Roger Tippett, Dr. Ian McCarthy, Rona Brennan, Vivien Cameron, Ishbel

McColl, Caroline Askew, Peter Willmott, Chris Cutts, Nicola Bissett,

Alistair Duguid, Alan Campbell, Alan Grant, Deborah Hamilton, Dave

Stevens, Melaine Fletcher, Mat Cottam and post-grads and under-grads in

Glasgow to numerous to mention The help of technical staff at the main

campus was often sought and received and I am particularly grateful to,

Davy Watson, June Freel, Willlie Orr and Pat McLauglin. Alan McGregor

and his staff in the IBLS work shop deserve special thanks for

constructing my various flume designs and altering them to the point

where they would work!

I have contacted many academics during this study for advise and I am

grateful to them all. Dr. Jill Lancaster was especially generous not only

with her time but ideas too. Her help extended to the practical by getting

me access to the Edinburgh University library while I worked in

Edinburgh during the winter of 1998-99. Other assorted folk at the

Freshwater Biological Association have been intermittently pestered for

help with identification of invertebrates and they also gave of their time

generously. Dr. Marian Scott, Geoff Hancock and Dr. Nigel Willby of

Glasgow University gave free and helpful advice on statistics and

dipteran biology and freshwater ecology respectively.

Last but not least I would like to thank Judith Milne and her kin and my

own family: Mum, Dara and Paula for all their support both emotional and

financial, go raibh maith agaibh.

This study was carried out whilst in receipt of a University of Glasgow

scholarship funded by a Scottish Office Initiative on improved river

management.

General Introduction

Chapter 1:General Introduction

The main aim of this project was to improve the scientific knowledge base of benthic

invertebrate - environment interactions in streams, to assist the better design of river

rehabilitation and management schemes. The basic question to answer is whether or not

instream flow preferences o f benthic invertebrates can be identified. There is a large body

o f work addressing various aspects o f this question but it remains to a large extent

unanswered especially in a quantified manner. The thesis chapters outlined at the end of

the introduction, show a progression from field surveys to laboratory experiments, where

initial measurements made in the field are tested in the laboratory. Finally macrophytes,

which provide a velocity gradient are examined as macroinvertebrate habitat. Part of the

aim of this work is to follow the ecohydrological approach, that is the combining of

ecology and hydrology to better improve our understanding and management of

freshwaters. In its infancy, this discipline still suffers from a lack o f basic definitions

hampering work. The first sections of this general introduction cover the basic biology and

physical structure of rivers, partly for general information but also to clarify some of these

basic definitions, and paradigms as I perceive them. A more specific introduction to the

work follows the general sections.

General background

1.1 The biology of flowing water benthic macroinvertebrates

Macroinvertebrates are a practical grouping of freshwater organisms, simply defined as

invertebrates occurring in or on (or associated with) the substrate, and visible to the naked

eye. Meiofauna which are invisible to the naked eye are often the more species rich and


abundant o f the two categories o f benthic fauna but remain relatively neglected (Poff et al.

1993; Ward et al. 1998). Being wholly dependent on size, membership of the two groups

is not exclusive, a situation well illustrated by the Chironomidae which can occupy both

groupings during their larval stage. Constituting a number o f phyla the macro

invertebrates exhibit a range o f life histories. The majority of the phyla spend their entire

life cycles in the lotic system; Crustacea, Mollusca, Annelida and Platyhelminthes1}

The remaining major grouping, the Insecta spend only their juvenile stages in the system,

although among some groups there are exceptions; the Coleoptera and the Hemiptera. The

insects are the most extensively studied o f the benthic invertebrates and several texts are

exclusively devoted to them (Merrit & Cummins 1979; Williams & Feltmate 1994).

Most o f the insects are univoltine or bivoltine but some species can take two years to

mature. It is very rare for animals to take longer than two years to mature in lotic systems

(Williams & Feltmate 1994). Adult emergence for the Ephemeroptera and some members

of the other groups is famously synchronised but time spent on the wing can be highly

variable. In the Trichoptera some taxa can spend the entire summer on the wing waiting to

reproduce. From a number o f studies it clear is that growth rates are plastic, and may

reflect ambient temperature and food availability or other environmental variables (Petts &

Bickerton 1994; Webb & Walling 1993). As the fecundity of these animals is almost

1 Other phyla do occur in running freshwater but were not encountered during this study. A checklist of the north European taxa is given in (Fitter & Manuel 1994). For simplicity the taxonomic hierarchy used in that publication is used throughout the thesis and is not a reflection of the author's views on the taxonomic structure of the Arthropoda which remains unresolved and controversial (Brusca & Brusca 1990).

2 There are a number of exceptions among the Gastropoda: it has been stated that it is difficult to distinguish between aquatic and terrestrial forms. (Macan 1977). The sub-class Euthyneura are exclusively represented by the order Pulmonata in this study are air breathers and can exist in quite dry conditions. Of their members the species most frequently encountered in the study is Artcylus fluviadlis which does not need to breath air and is wholly aquatic (Clegg 1952)!

2


exclusively dependent on juvenile feeding - see below - environmental conditions during

the juvenile stage are of the utmost importance.

Few adult aquatic insects have been observed feeding but some Trichoptera and

Chironomidae adults have been observed sipping nectar, Homopteran honeydew and sugar

water in the wild and captivity (Armitage et al. 1995; Malicky 1981). The contribution to

the animals’ overall fecundity is likely to be slight as these food sources contain little more

than carbohydrates (Svensson 1972). Dragonflies feed throughout their adult lives, but

were not encountered in the work presented here (Hammond & Merrit 1983; Miller

1987). Some Plecoptera (Nemouridae) need to feed before they can lay eggs, but even in

these cases the majority o f the adult biomass must come from larval feeding (Hynes 1976).

Diptera are the major exception, with fecundity closely related to blood feeding across

some of the families. In general though, investment by invertebrate adults in individual

young is limited.

Oviposition strategy can have a strong influence on the distribution of the juvenile forms

on the substrate and may, in the case o f insects and depending on species, be a product of

parental habitat use rather than larval preference (Harrison & Hildrew 1998). This is more

likely in lentic systems than lotic where there are fewer modes of larval dispersal. The

Gastropoda encountered in this study produce their eggs in jellied masses on rocks and

other submerged substrata (Clegg 1952) and in the case of Potamopyrgus jenkinsi by

asexual means (Maitland 1990). Among the Annelida, egg laying also occurs; the eggs

being laid in capsules (Brinkhurst 1963). In the insects, parental care appears limited to

3


oviposition. Eggs are laid in the water under stones (Baetis) on the water surface

(Ephemerella ignita) or on bankside vegetation (some Trichoptera).

The numbers o f eggs produced by all members o f the aquatic taxa shows high degrees of

intra and interspecies variation (Macan 1963), hinting at a range o f reproductive

strategies. Sexual reproduction appears to be the norm among aquatic insects but asexual

reproduction in the Chironomidae and Ephemeroptera has been recorded (Armitage et al.

1995).

The only aquatic groups which do show some maternal care are the amphipod and isopod

crustaceans which brood their young (Clegg 1952). It has recently been shown that the

amphipod Crangonyx pseudogracilis also actively cares for their broods by flexing their

bodies which alters the microhabitat o f their brood pouches (Dick et al. 1998).

Although fundamental to our understanding o f lotic ecosystems, little detailed work has

been done on the dispersal o f aquatic insects during the adult phase. Work on the genetic

variability between populations o f the Trichoptera in Australia indicate that there can be a

much greater transfer o f genetic material over large geographic areas than had previously

been believed (Hughes et aL 1998)3. Aquatic insects occurring in a river stretch can be

viewed as members o f metapopulations, where the entire population may range in spatial

occupation from a single island to the entire globe. Aquatic insects are thereby, potentially

resilient to localised disturbance o f lotic systems as long as their particular habitat patch

continues to exist after the disturbance event and can be colonised by animals from

1 Research on Plecoptera showed distinct differences between streams suggesting that they do not disperse to the same extent as the Trichoptera of the other study (Hughes et al. 1999).

4


unaffected sites, e.g., if aquatic insects are subject to metapopoulation dynamics (Hanski

1994 & Levins 1969).

Dispersal between non-contiguous river systems is particularly difficult to assess for non

insect taxa which have no aerial phase. The rate o f dispersal o f invasive Crustacea

{Crangonyx pseudogracilis) and Mollusca (Potamopyrgus jenkinsi) in the UK has been

impressive, with both invaders now found throughout the country, less than one hundred

and twenty years after being first recorded (Maitland 19904, Dick et al. 1997). Whether

this was purely mediated by humans or reflects a natural ability to disperse between

systems is difficult to determine at present. In the case o f Crangonyx there is good

evidence that the animal has moved between different catchments via the canal network

although this is not always the case. It is likely that humans have influenced the dispersal,

in the case o f many fish species (Adams & Maitland 1998) and zebra mussels (Buchan &

Padilla 1999). Future genetic work species and systems not heavily influenced by humans

would help determine the degree o f isolation o f populations o f these groups.

Once hatched benthic invertebrates find themselves near the base o f the food chain, usually

as primary consumers or detritivores, although some taxa are predatory right from

hatching, e.g. the Tanypodinae. Depending on food particle size and feeding mechanisms

of benthic invertebrates, the taxa have been assigned to functional feeding groups

(Cummins 1973). .Initially derived for insects only, it is now applied to the entire benthos

(Moss 1988). The primary distinction is between herbivory, detritivory and camivory.

Ephemeroptera are viewed as mainly collector gatherers feeding on Fine Particulate

5


Organic Matter (FPOM) and scrapers, Plecoptera as predators or feeding on Coarse

Particulate Organic Matter (CPOM) shredders. Tipulidae and Chironomidae can be

shredders, collector gatherers (filterers) or predators. The Simuliidae, the other dipteran

family are filter-collectors (Hart & Latta 1986). The Trichoptera and Coleoptera occur in

almost all o f the categories (Williams & Feltmate 1994: after Cummins 1973). The

gastropods are viewed as scrapers, the Annelida as predators or deposit feeders and the

Amphipoda and Isopoda as scrapers or collectors. The degree to which these groupings

are exclusive is less certain. Gammarus pulex have scraping mouth parts but can use these

to predate other amphipods which in turn may reduce interspecific competition (Dick

1992; Dick et al. 1990).

The link between feeding groups, animal mobility and morphology is strong. It is clear that

to be a filterer being located in areas of high velocity and on stable substrate is useful and

requires special adaptations e.g. suckers or retreats from energy consuming flow

conditions. For collector gatherers, mobility is important and the animal must be able to

swim through the water or crawl through / across the substrate matrix. It suggests there is

a link between species (identified using morphological characteristics) and how they

exploit their physical habitat. The river bed is heterogeneous and it is postulated that these

animals occupy different physical niches (areas of stream bed) depending on their

functional feeding groups.

Some of the major benthic taxa o f interest in this study have the ability to be either

parasitic or commensal on other invertebrates, feeding groups not addressed in Cummin’s

1 As a human food source invasive crayfish species are excluded from the example as the influence of man in their dispersal is well recorded and affords little room to speculate about natural mechanisms.( \ pseudogracilis is still largely absent from the Scottish highlands.

6


classification, (Cummins, 1973). Chironomidae appear best suited to this role and are

usually commensal, but can be parasitic (Tokeshi 1993). There has been at least one

record o f a Trichopteran (Orthotrichia) larva parasitic on other Trichopteran pupae (Wells

1992). The Hirudinea are o f course the most famously parasitic freshwater group but in

the systems studied here some leech species are predators, feeding directly on

invertebrates (Elliot & Mann 1979). Although it is felt that flowing water somewhat

negates the impact of ecto and endo parasites they do occur on the freshwater benthos.

Hickin (1967) reviews the epizoites and epiphytes found on Trichopteran larvae and

includes a mite, Atturus scaber infesting Goer a pilosa and protozoa on a range of other

species. Disease in general, is an area that has received little attention but could have a

profound effect on aquatic benthos distribution.

So potentially, predation, food availability, disturbance and physical habitat structure

alteration can affect the instream distribution of invertebrates and are well researched

(Boulton et al. 1992; Crowl & Schnell 1990; Crowl et al. 1997; Dahl & Greenberg 1996;

Death 1996; Dudgeon 1991; Dudgeon 1993; Dudgeon & Chan 1992; Hansen et al. 1991).

Before proceeding further it is necessary to describe the physical habitat of lotic systems

indicating the limits and opportunities which they present to aquatic benthic invertebrates.

1.2 Lotic systems: Physical structure

This section gives a short review o f the physicochem ical nature and geom orphological

structu re o f lotic systems. The section on the geom orphological structure o f rivers and

their substrates focuses on the subject o f the thesis, instream flow conditions; the physical

factors examined as possible environm ental gradients for invertebrates.

7


1.2.1 Life, light, temperature & water chemistry

The basic requirements o f almost all life: water, light, oxygen, carbon dioxide and

nutrients are available in lotic systems (Hutchinson 1957; Hynes 1972). Light penetration

in rivers is normally limited by the turbidity o f the water and it is only in the lower reaches

of rivers, or in very large systems that depth becomes a limiting factor (Hynes 1972).

Turbidity is usually dependent on discharge mobilising small particulate matter, and high

turbidity is therefore more frequent in winter when the temperature (in temperate rivers)

is too cool for much plant growth and the number o f hours o f sunlight are few and less

intense. Dissolved atmospheric gases are rarely limiting in running waters which are

frequently completely saturated or super saturated with oxygen, nitrogen and carbon

dioxide. Freshwater insects appear to be dependent on this high level o f available oxygen.

Lowering the percentage o f dissolved oxygen even slightly can have significant effects on

the health of rheophilic amphipods, Trichoptera, Simuliidae, Ephemeroptera and

Plecoptera (Golubkov et al. 1992; Kiel & Frutiger 1997; Macan 1963; Nagell &

Larshammar 1981). For some Trichoptera their cases facilitate oxygen uptake (Williams et

al. 1987) possibly allowing them to live in areas with lower oxygen levels. The capacity to

tolerate lower oxygen levels is observed across most o f the groups and tends to be found

mainly among burrowers.

Rain water contains varying amounts o f dissolved elements. The nutrient content o f the

river water is primarily dependent on the underlying bedrock geology, modified (often

substantially) by catchment land use and other anthropogenic factors. Land use,

particularly intensive agriculture and urban centres, have a detrimental impact on the

8


chemical quality o f water which may have a direct effect on the resident invertebrates. In

Scotland, surface drift which, when glacially derived, can be of different chemical

composition to the underlying bed rock, also contributes to the water chemistry (Survey

1971; Survey 1977).

The osmotic potential o f freshwater can obviously be variable for the same reason nutrient

concentration is variable. Aquatic insects are know to be hypertonic relative to freshwater

and are capable o f withstanding the normal fluctuations in its osmotic potential. They are

not capable of withstanding the essential potential o f salt water concentrations and this is

one of the reasons cited as limiting the lotic benthos to rivers and making it non

contiguous with the marine system5. Also it is one of the many reasons cited why aquatic

invertebrates do not drift too far downstream.

Temperature in running freshwaters varies more rapidly than in standing waters.

Superimposed on seasonal changes are diurnal cycles. Surface water streams reflect mean

air temperature over their entire length although this may alter as one proceeds down a

valley. Spring melt o f snows (which would be likely to affect all the rivers studied in this

work) may have temperatures below that of the mean air temperature for significant

periods subsequent to melting. On average, the upper sources of a catchment system tend

to be cooler than further down. Temperature can have a profound effect on larval growth

and emergence in the aquatic insects, best studied o f these is probably Baetis and

temperature is likely to effect the development of non-insect taxa in a similar manner, see

5 It has been pointed out that of the freshwater insects some groups have members occurring in marine environments. Tire trichopteran Philanisus plebeius lives in intertidal rock pools but has an additional organ not present in freshwater species which helps retain water within its hypotonic body (Leader 1976). As such the argument that limited osmoregulatory systems prevented insect colonisation of salt water begin to weaken.

9


Williams & Feltmate (1994) for a review of the effects of temperature on aquatic insects.

Feeding is rarely limited by temperature according to (Cummins 1973), but it can effect

instream distribution (Guinand et al. 1994).

Lotic systems have all the ingredients for primary productivity to succeed. They also have

another factor which to a greater or lesser extent excludes the growth of instream

macrophytes. That is the constant movement of water which makes the rooting of plants

not only difficult but, if they do succeed in rooting, subject to removal by flooding. The

constant erosion of fine sediment also leaves little suitable rooting material, most usually

found in the lower reaches o f river systems with mosses on rocks as the only rooted plants

in the upper reaches The macrophytes which do occur in rivers are adapted to this

disturbed habitat and often grow in shapes suitable for minimising drag (Sand-Jensen &

Mebus 1996). Phytoplankton are also often only found in the lower reaches of rivers:

frequently a major source of primary productivity is algae encrusting on rocks.

Filamentous algae also grow attached to the substrate and under suitable conditions, a

mixture o f Chironomidae, diatoms and other Protista grow among their strands to form

mats termed 'Aufwuchs' (Flynes 1972). Allochthonous material accounts for a large

proportion of the energy entering the system and the amount is closely linked to the

structure of the drainage system.

1.2.2 Drainage networks and Channel Structure

The sim plest hierarchy is that o f stream size; those tributaries furthest upstream are

sm allest in width and, as they progress to the sea, they join forming increasingly larger

10


channels o f higher stream order. The idea o f ordering streams is that o f R.A. Horton

(Hynes 1972). The instream structure o f channels also changes with distance from their

source. Higher up in the catchments slopes tend to be steep, quickly shedding water,

creating an erosional environment dominated by large substrate elements which can form

‘step and pool’ sequences (Carling 1995). Further down the system, the rivers still contain

a lot o f energy but are now wider and begin to deposit and erode material in a sequential

manner (Carling 1995). This leads to the riffle-pool structure where when discharge is

high pools are scoured out and the material deposited further downstream forming an

extended lip called a riffle (Clifford 1993). During low discharge, fine substrates deposit in

the pool sections, but not to the same extent in the riffle areas. Although relatively stable,

the riffle-pool system is a constantly shifting dynamic habitat (Carling 1995). In gravel bed

rivers some stability is created by the formation o f an armoured layer where, through

successive minor increases in discharge, the bed becomes compacted and in some cases

the substrate elements align their long axis with the direction o f flow. The armoured layer

allows the persistence o f finer substrate below this top compacted layer which, if it was

not present, would be eroded. These middle reaches are the subject of the work done in

this study this thesis. There can also be a zone o f low permeability below the river bed

were the water is no longer saturated with oxygen (van't Woudt & Nicolle 1978).

The deposition of material sorts it into mixed aggregates o f different sized substrate

elements; cobble bars, riffles, pools etc. These are often viewed as microhabitats for

macrobenthos (Brown & Brussock 1991) and this is one o f the questions addressed in this


thesis. As the landscape flattens, energy in the river dissipates and its ability to carry

sediment becomes reduced, here fine sediments become deposited as the river meanders.

The discharge down a river is seasonal, reflecting precipitation within its catchment. Such

fluctuations that do occur are classed by their return period, once in one hundred, twenty,

ten years etc. Related to their intensity is their capacity to move substrate and alter the

channel; some rivers in Scotland frequently migrate across their flood plains as a result of

large, intermittent floods (Smith & Lyle 1994), e.g. the River Feshie.

1.2.3 Scale ofphysical processes, implications for ecology

Climate and topography are obviously not the only factors important in fashioning a river

system and to aid biologists understanding o f these processes and their biological context

they have been categorised in a hierarchy with a number o f spatial and temporally scales.

A biological hierarchy o f processes has also been identified and linked to this physical

hierarchy o f factors. The simplest method o f linking the two is where the temporal scale of

a physical process is similar to that o f a biological one e.g. at a scale of 107 years,

megaform processes such as plate tectonics, climate change and eustatic change create

drainage networks which are on the same temporal scale as regional species pulses and

evolutionary differentiation. Each level in the hierarchy influences that below it. There are

a number of external physical and geomorphological processes working over a range of

time intervals which, with the physical size o f the area effected, are designated as mega,

macro, meso, and microform processes. Macroform processes include flood plain change

and channel evolution affecting river segments and are on the same time scale as possibly

short term localised extinctions and variations in the available habitat. Meso form processes

12


are listed as the influence o f shear stress, sediment deposition and channel processes which

effect reach pool/riffle systems - microhabitats and work over similar time periods as

metapopulation and patch dynamics and probability refugia. Microform processes include

annual flow fluctuations, scour and deposition working on fine scale patches and the same

time / and physical scale as the continuous distribution o f invertebrates.

1.3 Lotic system s:ecological interpretation

Ecology is the study o f the abundance, diversity and distribution o f living organisms in the

environment (Begon et al. 1996). Factors intrinsic to the macrobenthos and the extrinsic

or environmental factors affecting their ecology were reviewed earlier in the introduction.

General theories which attempt to explain the mechanisms underlying the ecology of

organisms are numerous and some have been applied to lotic systems. Others have been

developed specifically for lotic systems and what follows is a review o f some o f these

theories.

The River Continuum Concept which integrates the changing structure o f the temperate

riverine environment along its length postulates that the middle reaches o f rivers support

the greatest range o f animals (Statzner & Higler 1985). Reaches near the river’s source

lack light and therefore depend on allochthonous material supporting mainly shredders and

their predators. The main food sources in the lower reaches are resident plankton and

large amounts of FPOM derived from upstream sources which favours collector species.

13


The middle reaches are intermediate in type between the other two and therefore support

the most diverse community.6

The assumption o f the previous paragraph was that animals have different habitat

preferences and when the habitat is diverse, the community is too; it assumes the animals

are occupying different niches. The Competitive Exclusion Principle states that ‘complete

competitors cannot co-exist’. So how do so many species live together without driving

one another to the point o f extinction? This question is addressed indirectly in this thesis

by attempting to show that the animals present are using different ranges o f the flow

gradients present; that there is niche differentiation along these physical hydraulic axes.

There are numerous models to choose from which attempt to explain species richness:

Crawley (1986) lists eight. Freshwater ecologists argue as to whether the community

structure is in a state o f dynamic equilibrium, and hence structured mainly by species

interactions (Minshall et al. 1985), or whether the system is constantly being disturbed by

physical forces and is in non-equilibrium flux thus allowing species richness to be

maintained at high levels (Tokeshi 1994). Giller & Malmquist (1998) point out that a more

pluralist approach is now being adopted by ecologists and although this is the case it is

informative to briefly reviews the merits o f the different approaches (see Williams &

Feltmate (1994) for a full review).

Some o f the biological models (e.g. Spatial heterogeneity and The Musical Chairs Models)

are dependent on the habitat being patchy while the non-equilibrium models depend on the

6 The River Continuum Concept is based on a number of studies looking at the diversity of the benthos along river systems; these works include a study on the River Endrick, of which the Blane Water is one of the study sites used here. The River Endrick another of the sites would be viewed as rithron dominated and although a middle order stream is more upland in nature.

14


natural disturbance (flooding) o f river systems. Both are particularly applicable to lotic

systems which are believed to be both patchy and disturbed.

It is known that benthic invertebrates frequently have aggregated and patchy distributions.

The link between physical patchiness and invertebrate distribution has been made for

individual species and communities distributions (Elliot 1977). Patchy distributions o f

invertebrates in lotic systems are reported at a number o f scales; between-stream, and at a

finer scale in stream patches (Badcock 1976; Evans & Norris 1997). Instream conditions

can cause patchy distributions o f invertebrates at a between-stream scale (Rutt et al.

1989), but o f interest in this study are in stream distributions o f invertebrates caused by

instream habitat patchiness. Riffles and pools have already been mentioned as patches but

patches can also occur on a finer physical scale. (Minshall 1984) gives a comprehensive

review o f insect - substratum relationships in which he cites references to show that

animals have preferences for particle size, particle mixture and particle density (Malmquist

& Otto 1987). It has also been recorded that animals prefer different aspects o f stones

(Whetmore et al. 1990). Velocity and depth are important variables with patchy spatial

distributions and can cause the distribution o f macrobenthos to mirror this patchiness;

some caddis avoid areas o f the stream bed where they would have to expend energy to

withstand shear stresses (Bacher & Waringer 1996).

In lotic systems, disturbance is thought to be a major factor, although what constitutes a

natural disturbance for lotic invertebrates is hard to say. They can be redistributed by flood

events but normally they recover quickly unless the flood is very severe (Koetsier & Bryan

1995; Matthaei et al. 1996). As invertebrates drift as a normal means of redistributing,

15


mortality seems likely to be rarely caused by it and sub-lethal impacts the more likely

result o f most disturbance events.

The theories mentioned above tend to concentrate on one aspect o f the environment

(either biotic or abiotic) as the major determinant o f community structure. In reality there

are interactions between major habitat characteristics e.g. disturbance events are

ameliorated by habitat patchiness (Lancaster & Hildrew 1993) and more patchy habitats

are possibly most resilient (Death 1996). Patch type can also differentially increase or

decrease the impact o f a disturbance event, e.g. taxa associated with sandy sections were

significantly reduced after logging disturbance, but those on rock covered substrate

increased (Gurtz & Wallace 1984)

Competition for patches can be influenced by disturbance which complicates models such

as the ‘Musical Chairs Model’ which does not take into account disturbance at all, e.g.

some sessile benthic invertebrate species have been shown not to prefer the most stable

patches (stones) as predicted but those o f intermediate stability, hence manipulating the

relative importance o f competition for space (Malmquist & Otto 1987). When predation is

factored in along with competition and physical factors the situation can get very complex

(Hart 1992).

As the natural situation is so complex there has been a general consensus in lotic ecology

that it is most important to describe the scale, both physical and temporal, at which

processes (and models) are most likely to operate rather than concentrating on testing

models alone (Hildrew & Giller 1994). The categorisation o f processes at different scales

was described in detail in section 1.2 and it has been shown that hydrological factors can

16


act in a scale dependent manner (Statzner & Higler 1986; LeRoy Poff 1996). Processes

not only include the maintenance o f species richness, but also the mechanisms that the

animals use under these different circumstances, e.g., life history strategy. Rivers and

streams can be considered as a habitat templates on which the animal’s ‘bauplane’ adapts

(Brusca & Brusca 1990; Southwood 1977). Associations between the disturbance

frequency and habitat heterogeneity o f a system and its fauna have been devised

(Townsend 1989). In the general discussion (Chapter 7) the position o f the three rivers

examined are discussed in relation to this classification system.

1.4 Specifics of study

In an ecological context, the aims o f this thesis were simple; to identify habitat patch

preferences for benthic invertebrates and identify the relative importance o f two scales, the

larger being reach scale and the smaller at the scale o f Surber (lamboum form) samples,

thereby covering habitat produced by both meso and microform processes. The project

concentrates on small rivers typical o f those occurring in Scotland, from highland to

lowland conditions: River Etive (Highlands), Blane Water (Central Belt) and the Duneaton

Water (Borders): see Chapter 2 Section 2.2 for full site descriptions. All three river sites

would be o f middle order, 3-5. The river reaches studied here are all middle reaches and

were expected to have a wide range o f functional groups represented, in keeping with the

River Continuum Concept. This study focuses on flow variables at what can be viewed as

ambient, rather than disturbing, spatey discharges, and does not examine the impact of

predators, disease or disturbance on the distribution o f the macrobenthos.

17


There are a number o f key mechanisms in ecology which could influence the identification

o f invertebrate habitat preferences and are addressed here before proceeding to the details

o f the study.

Firstly, a suitable model describing the state o f the community has yet to be derived,

making it difficult to decide whether the benthic invertebrate community was at

equilibrium or in the process o f recovering from disturbance in the rivers examined.

Existing models such as the intermediate disturbance hypothesis do not successfully

describe community structure (Malmquist & Otto 1987) although other workers (Death &

Winterboum 1995) support the theory o f dynamic equilibrium at least at the patch level.

So where possible, data on the long term flow variability o f each site are provided. It was

assumed that despite the constant redistribution o f benthic invertebrates by flood events

and invertebrate drift, optimal habitat patches will have higher numbers of animals due to

‘bottlenecks’ occurring in the more suitable habitats (Townsend 1980). It was therefore

expected that the detection o f habitat preferences of even very mobile benthic

invertebrates was feasible.

The structure o f this study and the choice of environmental variables focused on direct

effects and therefore covers only a small subset o f the potential interactions which can

occur, e.g. those between the animals and their physical habitat, in the form of flow

preferences it is unable to detect some o f these interactions. The results presented should

therefore be considered cautiously, especially as such key mechanisms as competitive

exclusion may function causing some taxa not to be encountered in their preferred flow

conditions if a more successful competitor has already monopolised them.


As mentioned earlier environmental conditions during the juvenile stage are highly

important for the fecundity of these animals. Taxa which feed as adults were infrequently

encountered in samples in this study, although some simulids and fully aquatic

ceratopogonids were collected, both of which require blood feasts to reproduce in a

maximal fashion (Williams & Feltmate 1994)7. In this study I have assumed that any

correlations between physical variables and the distribution o f benthic invertebrates

reflects preferences of the juvenile stages and are not a function o f adult habitat selection.

This is valid given the lack o f overhanging vegetation, which adults use as shelter, at all

sites. Vegetation such as this has been reported to be influence juvenile distribution on the

river bed indirectly as the vegetation attracts egg laying adults (Harrison, in press). Having

now outlined the assumptions upon which the work is based I can proceed to the details of

the study.

Crustacea, Insecta, Mollusca, Annelida and Platyhelminthes were all encountered in this

study with both exopterygota and endopterygota insects represented. The hemimetabolic

developers in this study are the Ephemeroptera and Plecoptera. The holometabolic taxa

represented were the Trichoptera, Coleoptera and Diptera.

As mentioned earlier the rivers were chosen to represent the range of available conditions

in the middle reaches of Scottish rivers. Excluding the similarities mentioned above the

rivers differ not only in hydrology but a number of other factors. Like the majority of

running waters found north of the Highland Boundary Fault, the River Etive is on resistant

rocks, and consequently nutrient poor. The geology in the Borders and Central Valley (the


location o f the other sites used in this study) is mixed and the majority o f streams would

be naturally o f intermediate nutrient level (McKirdy 1999; Werritty et al. 1994).

The steep gradients of many Scottish valleys can cause shading of the river bed reducing

its productivity. The River Etive would be the most strongly effected o f the three sites

examined. Filamentous and encrusting algae are believed to be the main primary producers

within the rivers examined in this study. The moss Fontinalis does occur in these rivers as

does the macrophyte Callitriche stagnalis, but neither are known to be major sources of

food for macrobenthos and were not found in the sections o f river used in this thesis for

the identification o f general flow preferences (There is evidence that on occasion several

common invertebrates will graze most soft macrophytes).

In Scotland, staining o f water by humic acids from peat is common, giving the water a

brown colour which reduces light penetration (Hutchinson 1957). Again the R. Etive is the

site most likely to be affected by this phenomena, although the other two rivers have some

peat in their upper catchments too. Underwater ice does occur in Scottish rivers and is

likely to be common in the River Etive and Duneaton Water, both o f which are at high

altitude.

When choosing potential sites, frazil ice or possibly the first soft deposits o f anchor ice

were observed on the bed o f the Upper Clyde at Abington, less than 20 miles from, and at

the same altitude as, the Duneaton Water, one o f the other sites examined in this study

(see Hynes (1972) for ice definitions)8. The role of this natural form of disturbance is not

known. It is clear however that there is a large amount o f between site variation in

8 Trichoptera, Plecoptera and Ephemeroptera collected from the river bed ice remained immobile until they defrosted in the laboratory. Upon defrosting they became active and appeared healthy.

20


physical processes. At the very start o f this chapter, I mention that studies often suffer

from unclear definitions o f the physical habitat. The first results chapter o f this thesis tests

the hypothesis that riffles and pools can be described visually in a consistent manner or

with Jowett’s rule (Jowett, 1993) which is based on ambient flow measurements, e.g., that

these habitat units have constant physical characteristics at the three rivers examined and

that the benthos exhibit preferences for these habitat units irrespective of all the other

physical variables which may influence their distribution.

Recent work suggests that hydraulic habitat may be extremely important. Studies on a

regulated river in Norway have shown that growth rates o f Baetis rhodani increased post

regulation by ten to twenty times, (Raddum & Fjellheim, 1993). It was suggested by the

authors that, the reduced flow, subsequent increase in summer temperature and retention

of organic matter contributed to the increased carrying capacity for this animal. The

animal’s life cycle was also one month faster. Increased discharge, on another regulated

Norwegian river, appeared to cause an overall reduction in biomass o f invertebrates

(Fjellheim et al. 1993). At this site, rheophilic species increased in biomass, but lentic

species’ biomass decreased. These result suggests that indirect effects of altered flow

parameters are important and can have both positive and negative, sub-lethal effects on

benthic invertebrates. By identifying hydraulic habitat preferences o f benthic invertebrates

it was hoped that this work would improve approaches to habitat creation in river

restoration and rehabilitation schemes. A potential mechanism underlying these

observations is that the physical flow habitat is patchy and that invertebrates find some

patches more suitable than others. Chapter three, the second results chapter, tests the

21


hypothesis that benthic invertebrate distribution is patchy at the surber sample scale and

dependent on flow variables. By using multivariate analysis I was able to test the amount

o f variation in the benthic invertebrate community structure explained by flow variables.

As mentioned earlier, Townsend (1980) has postulated that it is possible to detect benthic

invertebrate patch preferences. By inference, if the animals are choosing patches on the

basis o f flow conditions we should be able to detect this by increased abundances in their

preferred patches. Chapter 4 examines the responses o f individual species to flow

variables. There is a presumption that taxa will congregate in areas with suitable habitat

conditions, frequently leading to the animaPs abundance showing a unimodal response to

environmental variables, Jongman et al (1988). I test the responses o f individual taxa to

see if they exhibit a unimodal response. This also allows us to visualise the degree o f niche

overlap along these environmental variables at the different sites. As sampling effort was

not equal at all points along the environmental variables gradients -samples were taken

randomly - it is possible that any positive responses are an artefact o f the sampling regime.

This aspect is also investigated.

Chapter five presents the results o f flume experiments designed to detect the upper flow

preferences o f benthic invertebrates. The main aim here was to try and replicate the results

of the field data.

The final results chapter addresses the use o f an instream macrophyte species, C.

instagnalis by invertebrates. In particular the diversity o f flow conditions within the plant

is examined and related to the distribution o f invertebrates. The hypothesis tested is that

different sections of the plants support different benthic invertebrate assemblages and that

22


these are consistent between different plants at the same site. That the outside o f the

plants, where velocity is highest, supports a more limited fauna is explored.

1.5 Thesis outline

• Chapter 2 is the first o f three concentrating on the ecohydrology of three Scottish

rivers. It focuses on large scale habitat units; riffles, runs and pools. General

descriptions o f the hydrological habitat of reaches examined are given. Jowett’s rule

for objectively identifying riffles, runs and pools is assessed and the biological

implications discussed. The following two chapters focus on finer physical scales.

• Chapter 3 discusses the distribution o f benthic macro invertebrates in relation to

hydraulic environmental variables at a finer physical scale than the previous chapter.

Ordination analysis is used to find patterns in the macro-invertebrate distributions;

aggregations or associated clumps o f taxa. The structure o f the community is then

related to the environmental variables measured, the relative importance o f the

variables is discussed.

• Chapter 4 reports individual taxon response curves to the flow variables measured.

The applicability o f Gaussian response curves to such data is investigated as is the

importance o f availability o f the environmental variable on the response o f the taxa.

Plasticity of species responses is discussed as are the implications o f deriving habitat

simulation models from such data.

23


• Chapter 5 investigates the responses o f individual benthic invertebrates to high

velocities. Behavioural observations are presented and some data on morphometries of

Ecdyonurus.

• Chapter 6 compares the abundance and diversity o f benthic invertebrates within stands

o f Callitriche stagnalis to substrate without vegetation. The hypothesis that the

outside o f stands represents an extreme environment are examined by comparing the

fauna o f the outer part o f stands to that found in the middle and underneath. Data on

the evenness and abundance o f taxa is presented. It is suggested that plant architecture

and velocity combine to create a range o f stability and thus microhabitats.

• Chapter 7 contains a general synthesis o f all the results and their implications for

benthic macroinvertebrate ecology and discusses how each set o f results complements

one another.

24

Reach scale comparisons

Chapter 2: Hydraulic and invertebrate surveys of reaches in the Blane

Water, River Etive and Duneaton Water

2.1 Introduction

Fundamental to river rehabilitation is the ability of researchers to describe a river reach

in language understood by both engineers and ecologists. Engineers and

geomorphologists are increasingly required to understand ecological requirements

when designing flood alleviation schemes and other works. Although frequently

working on a smaller physical scale, ecologists have increasingly recognised the

potential o f hydrological studies and techniques to describe the world of benthic

invertebrates. In the next three chapters I follow this trend by examining the hydraulic

world of invertebrates using a combination of ecological and engineering techniques.

Hydrological techniques have been applied to ecological studies in a rather piecemeal

manner, which has tended to make comparison of the applicability of such techniques

difficult. Hence in the recent ecological literature there have been calls for a consistent

systematic approach to hydraulic surveys of invertebrate microhabitats (Davis &

Barmuta 1989; Carling 1992). Davis lists a standard hierarchy of hydraulic parameters

to use when surveying invertebrate habitat, which is used here in a modified form. The

upper echelon of this hierarchy, involves measures of entire reaches, and it proceeds

down the physical scale to measurements of flow around individual stones,

transcending the scales used by engineers and ecologists. Results collected by using the

entire hierarchy facilitate an interdisciplinary understanding and allow comparisons

between work by different ecologists using the same methods.

25


The data presented in the next three chapters were collected over a survey period of

two months using the hierarchical approach mentioned above. The chapters are

structured on two themes. The first theme is physical and the second biological. This

chapter compares the three rivers investigated at the reach scale. The next chapter uses

data gathered on a smaller physical scale to compare community structure in relation

to flow parameters at the three sites. The final chapter concentrates on the responses

of individual taxa to flow, and possibilities for modelling these preferences. This layout

of results has the benefit o f making the transition from a larger physical scale familiar

to engineers, to the finer scale important to invertebrates.

In this chapter I attempt to differentiate between riffles, runs and pools (20m reaches)

using objective hydraulic data and individual taxon preferences. I also looked at the

same question subjectively by seeing if invertebrates identified elsewhere as having

preferences for either riffles or pools did so at our sites.

Riffles and pools are bed forms viewed by geomorphologists as possible primary

determinants of meanders and as discrete habitat units by ecologists, with runs

somewhat intermediate between the other two structures. In both disciplines an abiding

problem is the reliable identification o f these units. Investigators rarely give specific

criteria. Where criteria are given they tend to be relative visual estimates. Riffles are

frequently defined as steep, shallow reaches of fast, shallow flow with the water

surface broken by emergent substrate. Pools are deep slow reaches with an unbroken

surface, and runs are intermediate in form.

A number of techniques have been described which identify pools and riffles in a more

objective manner. Jowett (1993) has categorised these into those based on: substrate

size, water surface slope, the ranges of water depths and velocities, bed topography,

Froude number and water surface characteristics. Which criteria one uses depend on

26


the purpose o f the survey: geomorphologists tend to use changes in bed topography

and ecologists water depth to velocity ratios.

As objective measures, water surface slope and measures of longitudinal bed profiles

are the best. The main advantage of using changes in longitudinal bed slope, from a

geomorphologist’s point o f view is that it is independent o f discharge (unless discharge

is high enough to mobilise the bed). However this is the main disadvantage for

ecologists who are interested in studying aquatic invertebrates, as a riffle chosen using

this criterion may be dry when one goes to sample it, or its size may have altered

significantly from when it is first identified. A combination of water surface slope and

longitudinal changes in bed topography which can be assessed at the same time of

sampling, is most practical.

In this chapter I describe the survey reaches I studied using some of these criteria.

Following Jowett (1993) I also attempt to determine which physical parameters best

differentiate between riffles and pools.

As mentioned earlier, riffles and pools are viewed as distinguishable habitat units by

ecologists. This is most clearly evident in their use by fish, particularly by salmonids.

Invertebrate communities have been shown to differ between riffles and pools, with

riffles exhibiting greater species richness (Briggs 1948; Brown & Brussock 1991;

Egglishaw & Mackay 1966; Surber 1937; Wohl et al. 1995). Species composition

differs between riffles and pools reflecting the functional groups of the species present.

Higher numbers of collector gatherers, shredders and predators tend to be found in

pools (depositional areas). In riffles (erosional) these groups and filterers tend to be

represented too (Wohl et al. 1995) (NB Wohl does not include scrapers in his

assessment). The ability of heterogeneous habitats (riffle) to support greater species

richness is dealt with elsewhere but the capacity of riffles to support a more diverse

27


range o f functional groups than pools may be part of the explanation for this

phenomena. Care must be taken in interpreting these examples as contrary findings do

exist.

Armitage (1976) reported that on the River Tees, species richness is not highest in

riffles but is highest in the pools. This river was organically enriched which may have

lead to the high numbers o f Mollusca, Hydra and Nias sp. which the author attributes

to the high diversity in Tees pools. The velocities encountered in Tees riffles were

notably high, 0.5-0.75 ms'1 which may explain why they did not have the highest

diversity of invertebrates.

Biomass of invertebrates, (Wohl et al. 1993) a factor not addressed in this thesis, can

be greatest in either riffles (Briggs 1948; Brown & Brussock 1991; Surber 1937) or

other less heterogeneous areas including pools (Armitage 1976; Egglishaw & Mackay

1966; Hynes et al. 1976).

A suite o f other factors interact with the depth velocities typical of riffles and pools to

make these habitats selectively attractive to certain groups. As pointed out by Brown

(1987) adult riffle beetles (Elmidae) probably require water highly saturated with

oxygen for their breathing apparatus to function; they use an air bubble as a plastron

which absorbs oxygen from the water as it is used up the animal in respiration by

diffusion. The need for highly saturated oxygen conditions is also thought to be

important in the distribution of Plecoptera and Ephemeroptera and has been suggested

as a possible reason for their perceived preference for riffles (Nagell & Larshammar

1981).In the study described here I examine the distribution of laminar flow between

riffles and pools. Laminar flow conditions, which have the potential to limit oxygen

diffusion in rivers can occur in two different flow structures. Firstly the entire

boundary layer can be laminar or alternatively laminar sublayers to a turbulent

28


boundary layer are formed near the river bed. Whether laminar flow occurs is

dependent on the velocity of the river and bed roughness. If water flow is laminar,

diffusion o f dissolved gases and excreta is limited to a molecular rate, potentially

leading to toxic build up or (hypo) anoxia in the vicinity of an animal. In turbulent

water the random motion of molecules increases the diffusion rates preventing such

adverse conditions (Moog & Jirka 1999). Some caddis can use their cases to increase

flow rates over their gills and this could potentially provide a mechanism for dealing

with laminar flows (Williams et al. 1987). Under a fully developed turbulent boundary

layer a laminar sublayer may be present. Previously it has been suggested that this may

provide a refuge from turbulent flow for benthic invertebrates (Ambuhl 1959). This

theory was based on flume studies which over - estimated the depth of the sublayer

and was refuted by work using more accurate flow measurement devices; laser doppler

anometry (Statzner & Holm 1982) However this study was also based on flume work

and to date no field results on the existence or otherwise of the laminar sublayer at

invertebrate sampling points have been published.

This chapter has a number of aims which if condensed into one, would be the clear

description of hydraulic habitat at the three rivers examined at a riffle-pool scale and

the determination of any evidence that invertebrates distinguish between these habitat

units. Specifically the first aim is to test Jowett’s rule for differentiating between pools,

riffles and runs and comparing the results of it to visual categorisation, (Jowett 1993).

This rule uses cut of values of dimensionless combinations of flow variables, e.g.

Froudes number and a velocity-depth ratio. I attempt to see if the same cut of points

work at my sites.

The second aim is to test the hypothesis that invertebrate community composition is

not statistically different in the deep and shallow examined.

29


The final aim is to report estimates of near bed flow conditions and by so doing show

that, in the field laminar sublayers to the boundary layer are rare. The chapter also

contains a general description of the river sites examined.

2.2 Methods

2.2.1 Site choice

In an effort to cover the range of stream types found in Scotland a lowland river, an

upland river (rithron in nature) and a river in agricultural land with a flow regime

intermediate in character between the other two rivers, were sought for the study. A

number of potential rivers were screened using the following criteria:

• Non-channelled reaches o f river were available with sample reaches devoid of

macrophytes. (The effect o f macrophytes on flow and therefore indirectly on

invertebrate distribution is covered in a later chapter.)

• High Water Quality - Scottish Environment Protection Agency (SEPA) biological

and chemical water quality monitoring showed the rivers selected had high water

quality and a diverse invertebrate community. Where possible SEPA could also

provide long-term flow data.

• Accessibility - Each site was within 2-3 hours drive of Glasgow. All sites were

accessible by road and riparian owners were willing to allow sampling.

• Presence of pools and riffles - Suitable reaches were present which could be visually

categorised as riffles or pools.

Prior to the commencement sampling a number of rivers in each category were

assessed in this manner. Candidate sites were chosen and searched for suitable reaches.

Where reaches did suit qualitative samples o f invertebrates were taken and identified to

30


genus. Candidate intermediate rivers included the Mouse Water, North Medwin, the

Upper Clyde, Douglas Water and Duneaton Water; candidate lowland rivers included

the River Endrick, Douglas Water (flowing into Loch Lomond), River Kelvin (upper

reaches), Bannock Burn and the Blane Water; candidate highland rivers included River

Etive and rivers flowing of Rannoch Moor and those flowing into Loch Rannoch. The

three rivers selected best met the criteria outlined above: these were the Blane Water,

River Etive and the Duneaton Water, see Fig 2-1.

River Etive

Blane Water

Duneaton Water

Figure 2-1 Map of Scotland showing the locations of the 3 rivers sampled.

2.2.2 Description of sites

The chosen rivers fall into three of the four categories defined by the UK River Habitat

Survey (RHS): Steep streams (River Etive), Mountain valley rivers (Duneaton Water)

and Small lowland rivers (Blane Water). The fourth category (chalk streams) is non

existent in Scotland and hence the survey was representative of major Scottish river

types, although limited to three sites.

31


The River Etive (NN 224 521 UK National Grid Reference) is an upland stream

located north of the Highland Boundary Fault. Its upper catchment is largely occupied

by Rannoch Moor, a blanket peat bog overlying igneous resistant rock. Many feeder

streams are acidic but the main channel is circumneutral (pH 6.96, spot measure). The

catchment is generally unproductive with a low specific conductivity (mean

conductivity 62 pS cm'1). The site was sampled between the 9 and 30 August 1996.

The river forms the southern border of the Glencoe Regional Park. SEPA figures

available for the river downstream of the sampling point at site NN 136462 show

average annual rainfall at 2.5 1 per annum, and average daily flow as 9.671 mV1. The

river's invertebrates have previously been studied as part of the extensive and excellent

survey of highland streams (n = 52) by the Freshwater Fisheries Laboratory at

Pitlochry in the 1960s (Morgan & Egglishaw 1966), at a site upstream of the study

area used here.

The Duneaton Water (NS 869 221) is located at the head of the R. Clyde catchment

and is a major tributary of that system. Geology in the catchment is mixed and includes

coal seams and a range of other marine sedimentary rock types . Land use is mainly

mixed farming with heathland in the upper catchment. A SEPA gauging station

downstream from the site recorded minimum flow as 0.356 mV1 and maximum flow

as 85.390 mV1 between 1993 and 1996. Standard annual rainfall was 1.33 1. The site

was sampled between the 17-19 July 1996.

The Blane Water (NS 507 852) is a tributary of the R. Endrick and lies some 25 km

north of Glasgow. The site is situated in an agricultural area, downstream from a trout

fishery, and the river is quite nutrient rich. Lower Red Sandstone dominates the

geology at the site and within the catchment with a small pocket of Upper Red

Sandstone north of the site at Killearn and calciferous sandstone and basaltic lavas

32


present in the top o f the catchment. Glacial clay covers the bedrock throughout. A full

discussion of the biology and geology of the Blane Water is given in Doughty &

Maitland (1994). The site was sampled between the 3 and 9 July 1996.

2.2.3 Field measurements

Two reaches were chosen at each site, one deep and one shallow, estimated visually to

be riffles and pools. Riffle reaches had steeper bed gradients than the adjacent pool

reach and the water surface was broken. At the Etive site a clear riffle-pool sequence

did not exist. The ‘riffle’ reach did however have a steeper bed gradient and was

shallower than the pool with a lower mean velocity; see Table 2-1.

Table 2-1, Channel characteristics at the station reaches.

Reach Mean(cm)

depth Width (m) Mean velocity (ms’1)

Mean slope

Etive A deep 37.0 23.9 0.096 0B shallow 15.7 24.6 0.291 0.01232

Duneaton A deep 22.1 9.0 0.072 0.00022B shallow 13.7 3.7 0.271 0.00567

Blane A deep 22.1 15.3 0.182 0.00455B shallow 7.6 8.1 0.350 0.01695

A hydrological survey o f each reach (station) was undertaken to quantify available

habitat. The method used was a version of Instream Flow Incremental Methodology

(IFIM) previously adapted successfully for estimating available invertebrate hydraulic

habitat (Jowett et al. 1991). At each reach flow measurements were taken across a

total of 10 transects, at lm intervals, 0.5 m from the banks, and at the banks (only

eight transects were sampled in each reach at the Blane Water; this was due to the flow

meter breaking down). Transects were marked out in a sequential manner, at 1 m

intervals, moving upstream. The flow measurements included mean water column

33


velocity (mwc) and depth, which were measured using a SENS A electromagnetic

velocity meter on a calibrated rod.

At three random sampling points, across each transect, invertebrate density and

additional flow variables were sampled. Flow variables included, substrate composition

and a velocity profile. Substrate type was visually estimated (% scale) using the

following particle size scale: sand (0.06-2 mm nominal diameter), fine gravel (2-10

mm), gravel (10-64 mm), cobble (64-256 mm), boulder (>256 mm) and bedrock (solid

rock surface) (Jowett et al. 1991). A velocity profile o f 10 points was also recorded. It

was not always possible to record velocity profiles as the water was too shallow or no

flow could be detected and water depth sampled was limited to 1.5 m.

Quantitative benthic invertebrate samples were taken using a Lambourn sampler (frame

size 1375 cm2) at 3 random points across each transect. The Lambourn sampler had a

heavy skirt attached around its base to prevent loss of animals. Invertebrate samples

were preserved immediately in 70% alcohol. An equal amount of effort was expended

on collecting each invertebrate sample. Large substrate particles were scrubbed using a

nailbrush, after which remaining substrate was disturbed for 2 minutes to a depth of

10cm. All samples were sorted on white trays and animals identified to the following

taxonomic levels: Annelida to subclass; Diptera to family, genus and species; all other

taxa to genus or species. As identification to species was not possible in all cases

animals were assigned to operational taxonomic units (OTU).

2.2.4 Estimation of hydraulic parameters from velocity profiles

Hydraulic parameters were calculated to give a hydrological description of the two

reaches. Froude and Reynolds number were calculated from depth and mean water

column velocity at each sampling point. Shear stress, shear velocity and Reynolds

34


roughness number were derived from lognormal plots o f velocity profiles, see

Appendix II for an example and below for definitions of parameters. These parameters

are thought to be useful descriptors of near bed flow for ecologists (Davis & Barmuta

1989; Young 1993). Davis & Barmuta (1989) was used as the source for the correct

equations and choice of parameters. More recent clarifications of definitions and

applications o f some of these equations exist and were applied when necessary

(Carling 1992; Young 1992) (Carling 1993; Lamouroux 1993). Appendix II assesses

the limitations o f the technique.

Only parameter estimates generated from profiles which were statistically significant

(P< 0.05) were used, see Appendix III. This was based on the premise that being

dependent on the log normal relationship the parameter estimates are only valid when

vthe relationship is shown to exist, e.g., where it is statistically significant (Dingman

1984, Carling 1992, Smith 1975). Where data taken close to the substrate produced a

lognormal profile, disturbed only by gross turbulence higher up in the water column, a

truncated version of the profile was used; that is the lognormal readings were used on

their own. Recent evidence suggests this may give better estimates o f shear stress in

the field (Biron et al. 1998). The data from non-significant profiles was not completely

discarded. In chapter 3 (Section 3.2.1) mean water column velocity was generated

from the profiles and used in the multivariate ordinations, e.g. these points were

included in the ‘larger dataset’.

Classification of Flow

Reynolds numbers were calculated for each point of each reach. Re is given by:

Re = (UxD)/v

35


Where U = mean velocity, D = depth and v= kinematic viscosity. Reynolds number

describes whether mean flow is laminar or turbulent. Froude number further

differentiates flow. Froude number is given by:

Fr = U / J g D

Where g = acceleration due to gravity 9.8mV1. When Fr < 1 flow is designated sub-

critical (tranquil).

Flow in rivers usually forms a fully developed boundary layer. When this is the case a

velocity profile measured up through the water column can be used to measure

hydraulic parameters (see Appendix II). When the velocity profile is plotted on a

lognormal graph, shear velocity (u*) and equivalent bed hydraulic roughness (k s ) can

be calculated. Shear velocity is inversely proportional to the gradient of the velocity

profile and z 0 (characteristic roughness length) can be calculated by extrapolating the

regression line to zero velocity. The variable z 0 can be taken as an estimate of k s .

These variables are useful in classifying near bed flow. Reynolds roughness number

(Re*) can be calculated from these two variables.

Re, = n*ks / v

Critical values of Re, delineate flow as to the presence of the laminar sublayer o f the

boundary layer. Where the laminar sublayer is present its thickness (8 ) can be

estimated as follows:

8 = 11.6v / w,

2.2.5 Statistical AnalysisMann Whitney U tests were used to examine differences in Reynolds number, mean

water column velocity and, depth measurements between deep and shallow reaches

36


(Sidney, 1956). Spearman rank correlations were performed on the variables to

identify relationships between them. Discriminant Analysis was used to select the

variable which best distinguished between the deep and slow reaches. Discriminant

analysis is a parametric test which requires the data to be normally distributed.

Square-root transformations were used to normalise data.

TWINSPAN was used to objectively separate deep and shallow reach invertebrate

assemblages in each river. Twinspan (Two Way Indicator Species Analysis, (Jongman

et al 1988) is a hierarchical divisive clustering technique which classifies sites and

species by constructing an ordered two-way table from the site by species matrix. A

CA axis is generated and then divided at its centre of gravity (centroid) (Jongman et al

1988). The two groups are then divided further. This dichotomous branching continues

until a pre-set number of cuts have been under taken. Eigenvalues are quoted for each

division and can be interpreted as P values.

TWINSPAN was chosen over the rival clustering techniques, because it not only

provides a cluster analysis but also presents results in a table arrangement, which was

particularly useful in this analysis. It is viewed as the best method for table

arrangement and is recommended for hierarchical classification too (Gauch 1982).

Gauch (1982) does not address the use of fuzzy-clustering which is useful with

datasets where variation is continuous and TWINSPAN would force hard partitions on

what is effectively non-discrete data (Equihua 1990). It works by allowing samples to

belong to more than one group. This technique has been successfully applied to

terrestrial invertebrate data (McCracken 1994). However the whole premise of this

chapter is that there is a clear difference between the invertebrate populations in the

37


deep and shallow sections of the rivers, a sample can only be in one or the other, not in

both. Therefore TWINSPAN remained as the preferred ordination method.

2.3 Results

2.3.1 General Conditions

As expected, the deep reaches were slow-flowing and the shallow reaches faster-

flowing, see Table 2-1. In the Blane water velocity (0.182 ms’1) was quite high in the

deep reach, being greater than half the velocity in the shallow reach (0.35 ms’1). At the

other two sites mean velocity in the deep reaches was a third of that in the shallow.

This is in agreement with SEPA data which shows the Endrick/Blane catchment at the

time of sampling (1.3 m3s’1) was below mean discharge (2.3 mV1 averaged over 1993-

98) for the sampling month, but that it had experienced an increase in discharge during

the first sampling days, see Figure 2-2. That is it was experiencing a small summer

spate. The Duneaton Water was below mean discharge (mean =1. 1 m3s'1 averaged

over 1993-98) for the sampling month (July) at the time of sampling, 0.5 m3s'1, see

Figure 2-3. No long-term flow data on the Etive were available.

38


4 8

4.4

4.0

3.6

3.25° 2 8

f 2-4C

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Date

Figure 2-2 Mean daily flow in inJs 1 (9am-9am) for the R. Endrick, of which the Blane is a tributary, July 1996. Data from the gauging station at Gaidrew (NS 485 866) courtesy of the Scottish Environment Protection Agency (SEPA). Sampling dates are in bold italic type.

T - ( N ( 9 ^ i n ( O N c o o ) O T - f \ j { o t i n ( D K Q D a > O T - ( \ j ( o t i n ( O N c o a ) o ^T - T - T - r - ^ T - ^ ^ ^ t - f N f N ( N ( N ( N ( N ( N ( N ( N ( N ( 0 ( ^

D ate

Figure 2-3 Mean daily flow in m3s 1 (9am-9am) for Duneaton Water July 1996. Data from the gauging station at Maidencots (NS 929 259) courtesy of the Scottish Environment Protection Agency (SEPA). Sampling dates are in bold italic type.

A s it w as suggested that rivers like the R E tive do not break dow n into a p ro p er riffle

and pool sequence (Fozzard , et a! 1994) it is im portant to dem onstra te w hether the

tw o reaches exam ined w ere in fact different M edian channel depth w as significantly

39


deeper in the pool than in the riffle area of the river (Mann Whitney U test W=

80152.0, P>0.0001 adjusted for ties, Pool N=261 Median = 31 cm Riffle N=233

Median = 15 cm).

2.3.2 Comparison of flow conditions at different sites

I wished to see if the magnitude of Reynolds number could be used objectively to show

a difference between the riffles and pools chosen. In the case of the Blane and

Duneaton Water, Reynolds number was significantly lower in the pools; Mann Whitney

U test (Blane P<0.005 W=7938, Riffle N= 86 Median =37900, Pool N= 79

Median=20680), (Duneaton Water W=9224.5 P<0.0001, Riffle N=65 Median =

37620, Pool N=117 Median = 4200). There was no significant difference in Reynolds

number between reaches examined in the Etive.

Previous studies have shown strong intercorrelations between the environmental

variables which I used in this work. Some in fact are combinations of each other

(Reynolds number and Froude number are both derived from velocity and depth

measurements). A sub-set of measurements (those where substrate was also measured)

were examined using Spearman Rank correlation; see Table 2-2. The same general

trends show across all sites although there are site-specific differences. At all sites

velocity was significantly and positively correlated with Froude and Reynolds numbers.

Table 2-1, Spearman rank correlation matrix of hydraulic variables. Values are combined from the deep and shallow reaches at each river. *P<0.05.

velocity depth substrate Fr (mwc)

River EtivedepthsubstrateFrRe

-0.070.31* -0.11 0.95* -0.36 0.76* 0.56'

0.330.18* 0.53

40


Duneaton Waterdepth 0.11substrate 0.6* -0.02Fr 0.95* -0.2 0.59*Re 0.88*

*0.54 0.51*

Blane Waterdepth 0.23substrate 0.07 -0.003Fr 0.76* -0.22 0.10Re 0.78* 0.75* 0.08

*

0.68

*0.34

41


Table 2-2 Classification success of discriminant models using velocity, depth, substrate, and Froude number. Analysis was performed on square-root transformed data from each river separately and then combined. All figures are percentage of correctly classified samples. N = 60 (Duneaton w ater, R. Etive), N = 48 Blane Water.

deep A shallow Total PB

River EtiveVelocity 85 76 80 <0.00001Depth 59 80 70 <0.001Substrate 51 66 59 <0.05Fr 88 80 84 <0.00001

DuneatonWaterVelocity 68 75 71 <0.001Depth 64 67 66 <0.01Substrate 84 67 75 <0.00001Fr 80 85 83 <0.00001Blane WaterDepth 73 68 71 <0.01Fr 78 78 78 <0.01CombinedVelocity 69 72 70 <0.00001Depth 54 72 64 <0.00001Substrate 72 62 67 <0.0001Fr 71 70 70 <0.00001

Depth also showed the same relationship with Reynolds number across all sites. Depth

only showed a significant and negative relationship with Froude number in the Etive.

Velocity showed a significantly positive relationship with average substrate size in the

Duneaton and Blane. Depth was negatively correlated with substrate at all sites but

only significantly so for the River Etive. Froude and Reynolds numbers were

significantly correlated with all other variables. Froude number and depth at the

Duneaton Water and depth and substrate at the Blane were exceptions, not being

42


significantly correlated with either Reynolds number or Froude number. Discriminant

analysis showed that all variables, with the exception of Reynolds number (it was not

significant), could correctly classify riffle and pool reaches. In this subset of the data

velocity and substrate showed no difference between riffle and pool reaches. This was

not the case with the larger data set, where Reynolds number and velocity were

different at these sites. Froude number consistently showed the highest percentages

when it came to differentiating between the riffles and pools (always >70%: see Table

2-2). A combined model using all variables other than Reynolds number could identify

higher percentages than any individual variable for the combined data set (77%).

However it was believed that this result could be misleading (due to the high inter

correlations between variables) and the model was discarded.

0.6

0.4

0.2

0.00.8 1.60.0 0.4 1.2

Velocity m/s

Figure 2-4. Classification of pool, run and riffle by velocity/depth (V/D) ratios of 1.24 and Froude number (Fr) of 0.18; using Jowett’s rule any point with Fr less than 0.18 or V/D less than 1.24 is classified as a pool. Plot showing the combined deep (open) and shallow (solid) points for the three rivers from the smaller dataset (in which substrate variables were also measured); R.Etive (squares), Duneaton Water (diamonds), and the Blane Water (circles). For points with Fr > 0.18 and V/D > 1.24 (those not from pools) those from reaches with a surface slope of less than or equal to 0.0099 are classified as runs and as riffles if the slope was > 0.0099.

43


I applied Jowett's rule for differentiating between riffles, pools and runs, see Figure 2-4

and Table 2-3. Based on the outcome the deep and the shallow reaches have been

assigned to the distinct categories, see Table 2-3. As can be seen from Figure 2-4 not

all sampling points from one reach were in the same category, so I allocated one of the

categories to each reach on the basis of which category predominated for the reach

e.g. which category had the greatest number of points.

Table 2-3 Categories reaches sampled were allocated to using Jowett’s rule

River deep shallow

River Etive Pool RiffleDuneaton Water Pool RunBlane Water Run Riffle

2.3.3 Substrate available in the reaches

Substrate type at the three sites reflected their ambient flow conditions. The R. Etive

was dominated by boulder, cobble and gravel (Figures 2-5, 2-8). The Duneaton Water

was intermediate between the other two rivers having a high proportion of cobble in

the faster reaches, but generally being dominated by gravel and sand (Figures 2-6, 2-9).

The Blane water had hardly any large substrate elements and was almost exclusively

dominated by gravel (Figures 2-7, 2-10). With the exception of the Blane, in the

shallow reaches, larger substrate elements were more dominant than in the deep

reaches. To a small extent, the mixture of available substrates was greatest in the deep

reaches.

44


Figure 2-5, River Etive: substrate composition of the deep reach (a). Fingrav

Fine gravel

90 -

80 -

70 -

5 60 --i=>

& 50 -

40 -o 30 -

20 -

10 -

0 - □ 1 = 1 _— I----------- 1------------1----------- 1----------- 1------------1----G rav e l S a n d C o b b le F in g ra v B o u ld e rB e d ro c k

Sdostrate type

Figure 2-6, Duneaton Water: substrate composition of the deep reach (a).

45


90 -

80 -

60 “

40 -

30 “

1----------1---------- 1----------1---------- 1----------TG rav e l S a n d C o b b le F in g ra v B e d ro c k B o u ld e r

Sitostratetype

Figure 2-7, Blane Water: substrate composition of the deep reach (a).

90 “

80 -

70 -

s 60 -

a 50 “

40 ~o 30 -

20 -

10 -

o -— I---------- 1------------ 1----------- 1----------- 1----------- 1—B o u ld e r G ra v e l C o b b le F in g ra v B e d ro c k S a n d

Substrate type

Figure 2-8, River Etive: substrate composition of the shallow reach (b).

46


90 -

80 -

70 -

j* 6 0 -

I 50 "£ 4 0 “O

30 “

20 -

10 -

0 - 1 1 1 1 1 1-----G ravel C o b b le S a n d B o u ld e r F in g ra v B e d ro c k

Sitostratetype

Figure 2-9, Duneaton Water: substrate composition of the shallow reach (b).

90

80

70

60

50

40

30

20

100

G ravel C o b b le S a n d F in g rav B e d ro c k B o u ld e r

Sitostratetype

Figure 2-10, Blane Water: substrate composition of the shallow reach (b).

Only two thirds of the 136 velocity profiles recorded at the three rivers could be

measured with sufficient accuracy to estimate the presence and depth of the laminar

47


sub-layer. In cases where it was impossible to measure velocity profiles accurately this

was due to the gross turbulence of the water and it is reasonable to assume that a

laminar sub-layer would be prevented from forming in such conditions. Laminar sub

layers were rarely present elsewhere and when present were <lmm thick (Table 2-4).

Table 2-4, Characteristics of the laminar sub-layer of the boundary layer.

SITE N Depth of Laminar sublayer

SmoothPresent

TransitionalDisrupted

RoughAbsent

Etive 27 0.4 mm 8 12 15Duneaton 34 0.9 mm 8 20 9Blane 36 1 mm 3 11 17

2.3.4 Distribution of Turbulent and Laminar flow within reaches

Water flow at all sites was predominately turbulent or transitional, see Figures 2-11 to

2-13. Where flow was laminar it was in shallow water, near the channel margins. Flow

was also predominately sub-critical (tranquil) as determined from critical values of

Froude number.

48


10

10

10 '

Turbulent

10 '

10 '

Lam inar

1 0 ' 10 ' 10 10 '

Velocity M/s

10 10 10 '

Figure 2-11. Variation in turbulent flow in the River Etive. O denote the shallow reach and + denote the deep reacl^ The parallel lines demarcate the area of transitional flow Re = 500-2000 (Smith 1975).

Turbulent

ooQ .O)O + +

1 0 '1

+ +oocol + +

o oLaminar oo oo

10"1Velocity M/s

Figure 2-12. Variation in turbulent flow in the Duneaton Water. O denote the shallow reach and + denote the deep reach. The parallel lines demarcate the area of transitional flow Re = 500-2000 (Smith 1975).

49


Turbulent

o o

Q .(0o

10 '1

+ t+

+ oLam inar

10"1Velocity M/s

Figure 2-13. Variation in turbulent flow in the Blane Water. O denote the shallow reach and + denote the deep reach. The parallel lines demarcate the area of transitional flow Re = 500-2000 (Smith 1975).

2.3.5 Species lists

A comparison between the species composition of the reaches was performed on a

river by river basis. Table 2-6 gives the taxa list for all sites. A comparison of the

invertebrate assemblages in the deep and shallow reach of each river was performed

using TWINSPAN analysis, but it did not separate them clearly; data shown in

Appendix IV. At both the Blane and the Duneaton Water either the Chironomidae or

Ephemeroptera were numerically dominant. In the R. Etive, Trichoptera (mainly

Hydroptila sp.) were the most dominant with Ephmeroptera and Chironomidae the

next most dominant groups.

50


Table 2-5. List of taxa present in the deep (A) and shallow (B) reaches of the three rivers. Some taxa were identified to different levels depending on the quality of specimens. Chironomidae from the R. Etive were identified to genus and are presented in a second table. Names standardised to those in (Maitland 1977).

Operational Taxonomic Units OTU Etive Blane Duneaton

code

A B A B A BOligochaeta Olig 48 10 26 22Lumbricidae Lum 2 1Hirudinea Hir 8 4

Succinea Sue 3 1Ancvlus fluviatilis Muller Anfl 14 40 12 5Potamopvrgns ienkinsi (Smith) Poie 1 1

Hvdracarina Hvdra 7 1

Gammarus pulex (L.) Ganu 19 72 3 4Asellus aquaticus (L.) Asaq 6 18

Limnius volckmari (Panzer) Livo 23 29 41 71 8 45Oulimnius Ouli 0 2 35 48Oulimnius tuberculatus Outu 27 8Oulimnius troglodytes Outr 0 1Esolus parallelepipedus Espa 5 3Elmis aenea (Muller) Elan 19 44Dvtiscidae Dvti 2 0Dvtiscidae Oreodvtes? Dvor 5 1

Chironomidae Chir >1000 >1000 527 129Tanvnodinae Tanv 10 9Orthocladiinae Ortho 13 31Diamesinae Dia 0 2Simuliidae Simu 0 3 0 75Tipulidae Tinu 13 6 53 100

Baetis rhodani (Pictet) Barh 12 24 4 43 8 64Baetis scambus Eaton group Base 3 9Caenis rivulorum Eaton Cari 1 14C. luctuosa (Burmeister) Calu 0 1Ephemerella ignita (Poda) Enig 23 16 526 597 107 185Ecdvonurus Eaton Ecdv 0 2 67 75 86 265Rhithrogena semicolorata Rise 1 4Heptagenia Walsh Hept 2 7Ephemera danica Muller Enda 0 1Siphlonurus armatus Eaton ? Siar 0 1

Leuctridae Genus? Lege 77 99 3 5Leuctra moselvi Morton Lemo 3 2 28 92Leuctra fusca (L J Lefu 2 49Perlodes microcephala (Pictet) Pemi 0 2 0 1Perla bipunctata Pictet Pebi 0 1

51


Operational Taxonomic Units OTU

code

Etive Blane Duneaton

A B A B A B

Hvdrovtila so. Dalman Hvdr 86 21Oxvethira sp. Oxvt 16 1Polvcentropus kingi McLachlan Poki 3 4Polvcentropus flavomaculatus Pofl 8 5 1 0Plectrocnemia conspersa Pico 1 3Rhvacophila septentrionis Rhse 1 2Rhvacophila dorsalis (Curtis) Rhdo 1 3 0 3Agravlea multipunctata Curtis Agmu 0 1Hvdropsvche so. Hvdro 0 10Hvdropsyche siltalai Hvsi 3 6Rhvacophilidae Agapetus Rhag 0 4Glossosoma boltoni Curtis Glbo 2 66Lepidostoma hirtum (Fabricius) Lehi 3 0Anabolia nervosa Curtis Anne 9 0Agapetus fuscipes Curtis Agfu 1 11Psvchomvia pusilla (Tabricius) Psou 7 0Number of taxa 18 25 18 17 21 26Number of individuals 266 202 829 1196 930 1087

Table 2-7. List of Chironomidae present in the slow (A) and fast (B) reaches of the River Etive.

Taxa OTU code A B

TanypodinaeAblabesmyia annulata group 2 0Paramerina 5 2Trissopelopia 2 6Nilotanypus 0 1Macropelopia 1 0OrthocladiinaeCricotopus 7 21Heterotany tarsus 1 0Thienemanniella 2 0Tventenia 2 10Orthocladiinae Genus II 1 0ChironominaeVirgatany tarsus 0 1Tanytarsus 6 4Rheotany tarsus 5 1Total 11 6

52


Reach preferences of taxa

Correlation between deep and shallow section community structure was tested by the

non-parametric (G)amma Test on a contingency table based on table 2-6 . For the R.

Etive communities correlation was moderate G = 0.59, P < 0.0001, at the Duneaton

Water the correlation was weaker G = 0.45, P < 0.005 but at the Blane Water the

correlation was particularly strong, G = 0.67, P <0.001. Note however that for the

Blane Water the Chironomidae were excluded from the analysis as their numbers were

only estimates.

River Etive

Invertebrates from both the deep and shallow reaches were more typical of riffle fauna

than pool. Occurrence of animals was low, (468 specimens collected) reflecting the

poor nutrient status of the river. A total of 38 taxa was recorded from the two reaches,

29 taxa from the deep (pool) reach and 32 from the shallow (riffle) (note

Chironomidae were identified to genus level where possible for Etive samples but not

for the other sites, increasing the number of taxa). A biodiversity index of similarity

was calculated for the two reaches PSc = 52% (Cowan & Peckarsky 1990). This

indicates only a medium degree of similarity between the two reaches. OTU were

tested for differences in distribution between the two reaches using the Mann-Whitney

U test. Only two OTU showed a statistically significant preference for either reach.

Median abundance of Hydroptila spp was greatest in the pool reach, (W = 860.0,

P<0.001 adjusted for ties) while Caenis rivulorum median abundance was greatest in

the riffle reach (W = 610, P<0.05 adjusted for ties).

53


Duneaton Water

In total 2017 animals were collected. A total of 47 taxa was recorded from the two

reaches, 21 taxa from the deep reach and 26 from the shallow. OTUs were tested for

differences in frequency distribution between the two reaches using the Mann-Whitney

U test. The abundance o f benthic invertebrates was compared between the shallow

(run) and deep (pool) using the Mann Whitney U test. The test was applicable to 19

OTU (other OTUs had extremely low abundances and were excluded). Ecdyonurus

had a median preference for riffles (P<0.001). Other Ephemeroptera, Baetis rhodani

and Ephemerella ignita also showed a strong preference for the run reach (P<0.05).

Both species of Leuctridae showed a preference for the run reach as did the riffle

beetle larvae Limnius volkmari (P<0.05). The Tipulidae also showed a strong

preference for the run reach (P<0.05). The only OTU with a significant preference for

the pool was Oulimnius tuberculatus (P<0.05).

Blane Water

Excluding chironomids, a total o f 2025 animals was collected. In both reaches

chironomids were estimated at greater than 1000 individuals and not identified further.

A total o f 35 taxa was recorded fr om the two reaches, 18 taxa from the deep (run)

reach and 17 from the shallow (riffle). OTUs were tested for differences in distribution

between the two reaches using the Mann-Whitney U test. The abundance o f benthic

invertebrates was compared betweem the riffle and run using the Mann Whitney U test.

Other OTUs had extremely low abundances and were excluded. Baetis rhodani and

Glossoma boltoni showed a strong preference for the run reach (P<0.05).

2.4 Discussion

54


This study aimed to relate the physical attributes of deep and shallow reaches to a

subjective assessment of their habitat type. To confirm actual habitat distinctiveness

between reaches, I assessed the biological differences between the reaches. It was

possible to differentiate between sites using physical variables. Some species did show

preferences for either deep or shallow reaches, reflecting observations previously made

in the literature.

2.4.1 Physical character of the ReachesAt all three sites it was possible to differentiate between riffle, run or pool using the

range of depth and velocity in the reaches. Froude number was the best indicator of all.

Jowett's rule did show some of the reaches to be runs which was not expected and did

not conform to the visual assessment of the reaches; I chose the deep reaches as pools

and shallow reaches as riffles. Noteworthy was the ability of Jowett's rule to

differentiate between the deep and shallow reaches of the River Etive. As mentioned

earlier the river is rithron dominated and does not have a classic riffle pool sequence.

The sites on which Jowett based his work included boulder strewn reaches, which may

have been similar to the R. Etive (Jowett 1993). Alternatively this may be the result o f

a sampling error on my part as the pool and riffle could not be differentiated using the

larger data set which included sampling points which had non significant velocity

profiles.

There are consistent differences between all the reaches which suggest that the suite of

physical variables measured did differentiate between them. With the exception of the

Blane, the sites showed the deeper reaches to be depositional and the shallower

reaches to be relatively erosional. This pattern is consistent with the definition of pools

as depositional zones and riffles as erosional zones. The fact that the deep reach of the

Blane does not follow this pattern would suggest that it is a run. The water in both


reaches o f each river was predominately turbulent with laminar flow being rare. This is

also consistent with the findings of other studies and the predictions of

geomorphologists.

Others using a more descriptive classification of near bed flows have suggested that

the habitat classifications o f pool, riffle and run do not necessarily apply to near bed

habitats as all their near bed flow categories were recorded in each class (Young

1993). It is perhaps better to view these habitat classes as aggregates of smaller habitat

units especially as they do have some biological significance, see below.

2.4.2 Biological character of the Reaches Diversity

The riffle reaches supported greater numbers o f taxa than the deep reaches conforming

with the norm in other studies, see introduction for references (reach 1.1. paragraph

9); however again the Blane was an exception.

Species composition

Ephemeroptera and Chironomidae dominated the fauna at all sites a situation noted in

previous comprehensive surveys o f the summer fauna of highland rivers, (Morgan &

Egglishaw 1966). The methods employed by Morgan and Egglishaw were similar to

the one I used, being a combination of disturbing sediment and stone scrubbing adding

a degree of certainty to the comparison. This result reflects the seasonality in relative

abundance of taxa as the same survey found the spring fauna to be dominated by

Plecoptera and Ephemeroptera and suggests applying data gathered on benthic

commmunity structure and composition in one season to other seasons is potentially

misleading.

Hydroptilidae were detected in the River Etive in high numbers but not at the other

sites. This may again reflect a seasonal effect as the R. Etive was sampled later in the

56


summer, in August, than the other two sites. This hypothesis is supported by a

seasonal study on invertebrate fauna in another Scottish river, Shelligan Bum a

tributary of the River Almond, where Hydroptilidae were detected in low numbers

until August when their numbers increased (Egglishaw & Mackay 1966). As the taxon

is known to occur in both the Blane Water and the Duneaton Water (SEPA pers

comm.) it would seem likely that the effect is seasonal rather than the R. Etive being a

site o f particular suitability for the taxon. Further evidence is that during their sampling

of the R. Etive Morgan & Egglishaw (1966) found the taxon in low numbers; their

sampling was in July, the same period the my sites were sampled.

Caseless caddis were represented by more species in all rivers than cased; the Blane did

have the highest number of cased species. Again this mirrors the findings of other

workers (Armitage & Gunn 1996; Morgan & Egglishaw 1966) who also found as I did

that the Orthocladiinae dominated the chironomid fauna. The Etive was the most

extreme in this respect with the largest number of caseless caddis species represented

of all the rivers and probably reflects the instability of depositional habitat within the

site; the larger boulders and lack of small substrate elements at the site attest to this.

Cased caddis do tend to require more stable conditions on the whole. Both the Blane

and the Duneaton had greater amounts of finer substrate, no or few boulders and

greater numbers of cased caddis species than the Etive. Using riffles, pools and rock

outcrops as mesoscale habitat units others have identified functionally distinct caddisfly

communities (Huryn & Wallace 1988). There is some evidence from this study that this

might be occurring here too. The Hydropsyche larvae preferred the fast reaches,

Anabolia nervosa and Glossoma boltoni favoured the slower reaches and there is the

possible link between Hydroptila spp. and boulders. Some of these relationships are

57


not statistically significant and the functional feeding strategies of some of the species

need to be clarified.

Other unique aspects of the R. Etive species assemblage is the absence of both

Mollusca and crustacean macroinvertebrates which is likely to be a product of the

underlying geology as is the case with the high numbers of Mollusca in the Blane

Water (Egglishaw & Morgan 1965). Macan (1977) divides the British freshwater

molluscs into two major groups based on their preference for hard or soft waters.

Calcium which is needed to form shells is the underlying determinant for separating the

two groups and is an element which is likely to be in short supply at the R. Etive.

Individual species responses in relation to site specific factors

A number of species showed preferences for either deep or shallow reaches,

suggesting that for some invertebrates at least, the different types of reaches represent

habitat units. How Hydroptila species avoid getting washed out of the Etive system is

interesting to ponder; it may be that their small size allows them to seek refuge in the

filamentous algae on which they are known to feed. I observed in the field that the

greatest amounts of filamentous algae were on boulder tops, submerged in the deep

reach, a most stable substrate. Alternatively the preference of Hydroptila species for

the deep reach for the R.Etive may be explained by the need of this taxon for very fine

sand grains to form cases during it's last instar (Wallace 1981). Sand did occur with

greater frequency in the deep reach of the river. Hydroptilidae also had a significant

preference for pools in the Almond Water (Egglishaw & Mackay 1966). Caenis

rivulorum the only other animal found to have a preference in the R.Etive was found

by Armitage (1976) to prefer the 'intermediate' (0.2-0.6 ms'1) reaches of the R. Tees,

which has comparable velocities to the Etive riffle reach.

58


The preferences exhibited by both mayfly and stonefly in the Duneaton water (run) for

faster reaches have been observed elsewhere in British rivers, in the Tees (riffle), the

Afon Hirnant (riffle) (Hynes et al. 1976) and the Almond (riffle), although the reflected

patterns were not perfect. In the Tees, Leuctra fusca and Ecdyonurus dispar were

among some of the few taxa to show no preferences, whereas in the Almond

Ecdyonurus spp. were in greatest abundances in the pools in April and the riffles in

September. Morgan & Egglishaw (1966) believed this seasonal variation was due to

different preferences being exhibited either by Ecdyonurus spp. instars or species.

Alternatively they believed that pre-emergence behaviour altered the animals habitat

choice. Again it is interesting to note that Limnius volckmari had a preference for the

pools in the Tees which were o f similar velocity range as the run in the Duneaton

water to which it was also partial (Rosillon 1988).

Contrary to its previously reported preferences I found Tipulidae to occur most

frequently in the riffle of the Blane which is difficult to explain. Easier to explain, yet

equally at odds with previous studies, is the preference exhibited by Baetis rhodani for

the run reach. An extensive study covering a greater number of rivers found Baetis

species indicative of riffles (Harper et al. 1998). This trend can be seen in the

Duneaton Water where Baetis rhodani also showed a preference for the run. The fact

that the animal occurred with greater frequency in the run of the Blane Water suggests

that it was avoiding the riffle reach o f the stream (the run reach was immediately

downstream of the riffle). Drift rates of baetid mayfly have been experimentally shown

to be greater during spates (Lancaster 1992). Harper's study (1998) also showed

Caenis species as being indicative of silted riffles and runs whereas my data suggests

they occur preferentially in the shallow reaches of the R. Etive.

59


Some OTUs showed differential preferences for the reach types present at different

sites. Based on substrate type the R. Etive experiences the highest shear forces of the

three rivers with the Duneaton Water having the next highest. As the substrate was

only sampled to a depth of 10cm, composition of the armoured layer only was

measured. Hence the substrate types reflect the rivers’ recent past only. Any larger

substrate elements will have been buried by more minor spates. In this context the R.

Etive is likely to have more intense spates than the other two rivers, as one would

expect from the steep gradients in the catchment. I suggest that the intensity of recent

disturbance at the three sites, as experienced by invertebrates is most intense in the R.

Etive, least intense in the Blane Water and intermediate at the Duneaton Water. This

measure does not differentiate between disturbance events with high frequency, low

force and low frequency, high intensity. There is also the possibility that the armoured

layer, if compacted by low intensity spates becomes more resistant to entrainment and

this index based solely on substrate type could underestimate the resistance of the

substrate. This is most likely at the Blane where the substrate appeared quite compact.

2.5. Conclusions

2.5.1 Physical study• Jowett's rule accurately differentiated between riffles and pools; it also identified the

deep reach of the Blane as a run highlighting the danger of choosing a site during

one set of flow conditions and sampling under another. It also classified the R. Etive

reaches into pool and riffle although geomorphologists would disagree on the

grounds of the highly erosional nature of the stream and ecologists on the basis of

the predominantly riffle type fauna of the ‘pool’. In rivers of high slope the rule

should probably be applied with caution if at all.

60


• Separating pools, runs and riffles on a subjective observational basis is insufficient

on the evidence of this study. Some measurements of velocity and depth need to be

performed.

• A large proportion of the physical variability between the riffles and pools was

consistent between sites. Further the mixture of substrate in the deep and shallow

reaches suggested that they were depositional and erosional zones respectively.

• Flow was turbulent in all reaches with a measurable laminar sub-layer being rare. I

conclude that at the substrate surface oxygen is unlikely to be a limiting factor.

2.5.2 Biological study• Each river had a unique assemblage of invertebrates, although there was a large

degree of overlap in composition between sites. Differences could be explained by

comparing my results with previous studies on the distribution o f invertebrates in

relation to geology and water chemistry.

• Riffles supported a greater diversity of species than the pools although the

assemblage of species in both pools and riffles were not sufficiently different to

separate them using TWINSPAN.

• Some species showed preferences for either deep or shallow reaches, these

preferences were usually site specific and reflected a large number o f local factors,

but velocity and substrate type are probably the most important.

• Based on the physical and biological data it seems clear that riffles and pools

represent local combinations o f physical habitat conditions which alter temporally

and are fundamental to maintaining the diversity of macroinvertebrates in Scottish

rivers.

61


• Most taxa can survive outwith their preferred habitat unit (either riffle or pool)

suggesting that their own physical habitat preferences are acting on a smaller

physical scale.

• The species that make up the assemblages are likely to exhibit preferences for the

environmental variables associated with pools and riffles e.g. depth, velocity,

substrate type etc. The evidence also suggests that these variables are fundamental

to niche separation among the species present. This is explored in the next two

chapters.

62

Community ordination

Chapter 3: Benthic invertebrate community ordination

3.1 Introduction

3.1.1 Outline

The purpose of this chapter is to examine the distribution of benthic invertebrates in

relation to hydraulic environmental variables at a finer physical scale than the previous

chapter. The data presented here reflect the habitat complexity on an intra-reach scale

and gives some indication of inter-river variation. The analysis consists of two stages:

the first attempts to find structures in the invertebrate community e.g. associations or

clumps o f species from samples within the same river. The second stage tries to relate

community organisation to environmental factors. By using ordination and cluster

analysis it was hoped to identify groups of invertebrate species with indicator species

which have similar flow requirements.

3.1.2 Physical patchiness

The design of a program monitoring the ecohydrological health of rivers could

successfully incorporate the results of such analysis. By exploring species distribution

independently of environmental variables in the first instance the two stage process has

the advantage, of following (Pringle et al. 1988) advice - 'that variability among

patches in lotic systems be viewed as valuable information rather than statistical

noise to be overcome by manipulating sampling protocols'.

Hildrew & Giller (1994) explored the link between habitat patchiness, scale,

disturbance and the stream benthos. They made the link between the concept of habitat

template (the river: Southwood 1988) and benthos, pointing out that although the river

63


habitat is heterogeneous, clumping or aggregation o f similar types of habitat is

common:

"Undoubtedly the major architects o f physical patchiness in streams are the forces o f

flow. Flow is scale-dependent, being heterogeneous and 1aggregated' (i.e. different

habitat "patches' have different average flow characteristics) at different spatial

scales ranging from the sub-millimetre to whole river reaches or entire drainage

networks. "

3.1.3 Spatial aggregations

The more obvious aggregations include riffles and pools (which were the subject o f the

previous chapter), but within these there occur smaller habitat units. In a diagram

illustrating scale effects (Hildrew & Giller 1994) showed fine gravel patches, moss on

boulders, transverse bar o f cobbles and sand and silt over cobbles as aggregate units at

the 'microhabit scale (1 0 1 m ). By using cluster and ordination methods it might be

possible to reflect this pattern, by clumping species together that occur in the same

habitat aggregates. Although the quadrat size used to sample invertebrates is on a

comparable scale (sample area = 0.0625 m2) to the microhabit scale, in my study there

was the possibility of sampling across two or more discrete habitat units, due to the

random choice of sampling points. All physical variables were measured on continuous

or integer scales so that taxa responses to variables could be observed even when more

than one microhabit was sampled.

3.1.4 Assessment of the ecohydrological health of rivers

Research in this area is driven by two forces, firstly the pure scientific interest in niche

separation and habitat patchiness in rivers and secondly the development of

conservation and monitoring strategies for river health using responses of benthic

64


invertebrates to flow in regulated rivers. The 1999 draft of the EC Water Frameworks

Directive (19 February 1999) underlines the potential importance of this type of study.

The draft requires member countries to identify high quality reference sites for surface

waters, including rivers (Annex II section 1.3, paragraph iv). Ecological status is to be

defined using biological and hydromorphological elements. The hydromorphological

elements are defined as those which support the biological element and include

hydrological regime, river continuity and morphological conditions; depth, width,

structure of bed and substrate type (Annex V section 1.1.1). Under the directive these

pristine rivers would form a standard all other rivers would have to reach. In

preparation for this legislation SEPA are currently trialing Instream Flow Incremental

Methodology (IFIM, (Bovee 1982)) and its associated computer programme Physical

Habitat Simulation (PHABSIM). PHABSIM is one of a number o f models which

evaluate riverine physical habitat, specifically it models the available habitat for aquatic

species at different discharges (Johnson & Law 1995). Its successful application

elsewhere as a means of estimating instream habitat make it a suitable candidate ( e.g.

(Armour & Taylor 1991) reports 616 applications o f IFIM up to the early 1990s in the

United States alone). Both the methodology and the program have mainly been applied

to salmonid fisheries management (Armour & Taylor 1991; Johnson et al. 1995;

Jowett 1998; Strevens 1999). This list o f references is not by any means exhaustive:

rather it is intended to give the reader a representation of the work being carried out on

different continents.

IFIM has been used to estimate habitat suitability curves for benthic invertebrates

(Jowett et al. 1991) and PHABSIM predictions o f increased habitat availability for

benthic invertebrates in artificial riffles have shown positive and significant correlations

65


with actual measures of increased biodiversity (Gore et al. 1998). By using a variant of

IFIM developed for collecting data on benthic invertebrates (Jowett et al. 1991) in the

work reported here, it is hoped to make the work comparable to other similar studies.

When using PHABSIM there can be a conflict between the flow requirements of

different species. One aspect o f this study is to attempt to find clusters or guilds of

species associated with combinations of flow variables which simplifies this

management problem. This technique has been successfully applied to the management

of fish stocks (Jowett & Richardson 1995).

Criticism can be levelled at PHABSIM on a number of fronts, probably the most

fundamental being that annual flows for a river are determined from one or few 'snap

shot' IFIM studies which only reflect the fauna's preferences at the time of sampling.

Another serious problem is that PHABSIM assumes 'that target organisms have

specific microhabitat preferences and the ability to move to areas o f suitable

hydraulic conditions in response to changes in stream discharge' (Layzer & Madison

1995). Although this is obviously the case for invertebrates which end up in flow

refugia during spates (Lancaster & Hildrew 1993b), as Layzer and Madison point out

these assumptions do not apply to sessile or slow moving invertebrates e.g. Mollusca.

(they also point out that their study organism, freshwater mussels, appears to have

different flow preferences at different discharges). Neither does PHABSIM address the

importance of disturbance flood events which redistribute and alter the species

composition of riverine invertebrate communities (Boulton et al. 1992; Brooks &

Boulton 1991; Cobb et al. 1992; Dudgeon 1993; Hildrew & Giller 1994).

The aims of this chapter are to test the structure of the benthic invertebrate community

at the surber sample scale, identify any aggregations in the distribution of taxa and

66


estimate the influence of flow variables on the variation in community structure using

multivariate analysis. The hypothesis is that the distribution of benthic invertebrates

should be clumped and that flow variables should describe a large amount of the

variation. The interrelations between the environmental variables are described in this

context.

By addressing these questions PHABSIM is also tested indirectly, as it assumes that

benthic invertebrates are responding to flow variables at this scale, in a predictable

manner.

3.2 Methods

3.2.1 Data Collection and Dataset Structures

Collection methods for data examined here are described in the previous chapter. For

each river data from both reaches were amalgamated into a matrix of taxa by sites, and

a second matrix containing environmental variable by site data. For each river the

ordination was attempted using an environmental variable data set which contained

variables derived from velocity profiles.

The profiles were measured at all points where invertebrates were sampled. For each

profile, velocity should show a linear relationship with log depth if a full developed

boundary layer exists and the layer is not disturbed by gross alterations in water

movement, e.g. a velocity profile taken downstream of a semi submerged boulder may

show a log linear relationship at the depths above the boulder but back edding may

distort the profile lower down. Not all profiles had significant R2 values and these were

discarded as near bed flow variables could not be calculated from them being

dependent on the log linear relationship between depth and velocity being intact, see

67


chapter 2.2.4 and Annexes II and III. It has been suggested that some data could be

salvaged from these profiles. The mean water column velocity was calculated for these

points from the profiles allowing their inclusion in the ‘larger dataset’, see below.

An alternative method of salvaging data from the profile was to use them to estimate

turbulence above the bed, using their deviation from the predicted log-normal

relationship as a turbulence surrogate. Given that turbulence at the bed is the factor of

most interest, as that is where the animals are and, not turbulence further up the water

column, the use of the profiles in this way was not suitable.

Two data sets for each river were now analysed. The dataset containing sampling

points for which velocity profiles were intact and a second larger dataset which

included all sampling points where invertebrates were sampled. The larger datasets

contained a more limited array of environmental variables; mean water column

velocity, depth, % sand, % fine gravel, % gravel, % cobble, % boulder and % bed

rock. The smaller datasets with significant velocity profiles also contained these

measures.

Where the smaller data sets proved unsuitable the larger parent data set was used

instead, see section 3.3.1. For the R. Etive where many taxa occurred in few samples

the data set excluded taxa occurring in less than 5 samples.

3.2.2 Analysis procedure

The production of an ordination analysis requires refinements to the input data which

influences the final result as does the structure of the data sets involved. Initially a

TWINSPAN analysis was carried out on the data from each site and the groups

overlaid in a Detrended Correspondence Analysis (DCA) taxa scatter plot. As

TWINSPAN and DC A both use the same algorithm to order a species by site matrix

this process is acceptable. This allows TWINSPAN groups to be visualised in an

68


ordination diagram as opposed to a dendogram aiding interpretation and allowing

comparison with constrained ordinations (see last paragraph for an expanded

explanation). No TWINSPAN analysis was performed on the Blane Water data

because DCA showed the length of the environmental gradient to be short suggesting

taxa were exhibiting linear responses to the environment. As TWINSPAN is based on

a unimodal model it was not applicable.

Outlined below are the methods used to derive the final ordinations, highlighting the

peculiarities o f each river's data set and the refinements made. For each site a DCA

was carried out initially to measure the length of the environmental gradient as

suggested by the CANOCO computer program help (ter Braak 1997, 1998).

Depending on the length o f the gradient, either a unimodal (gradient >3-4 s.d.) or

linear (gradient <3 s.d.) based ordination technique was used. The results o f each

ordination were interpreted with the aid of the explanations given in the CANOCO

manual, chapter 6 (ter Braak & Smilauer 1998) and adjusted where necessary.

Constrained ordination techniques (Canonical Correspondence Analysis (CCA),

Redundancy Analysis (RDA)) have been criticised for being potentially misleading

(McCune 1997), so some clarification of their purpose is useful. Other ordination

techniques which are not constrained describe community structure e.g. DCA scatter

plots will clump species together which co-occur. Constrained ordination techniques

do not: primarily they relate species distribution to environmental variables. If one

wishes to relate community structure to environmental variables one should apply an

indirect gradient analysis and subsequently relate this ordination to environmental

variables, e.g. DCA and superimpose the environmental variables on top of a scatter

diagram. Jongman et al. (1988) suggests a number of methods for relating

69


unconstrained ordinations to environmental variables. Another useful alternative is to

first carry out an unconstrained analysis and compare it to a constrained one as is the

case herev\Here I compared DCA to CCA, though there has been the criticism that

DCA should only be compared to DCCA, presumably on the basis that both are

detrended. Detrending is a remedial tool for fixing the ‘arch effect’ in ordinations

(Jongman et al, 1988). It was not necessary to apply it to the C C ^ here. A dataset

which produces an arch effect when processed by unconstrained ordination does not

necessarily do so when the ordination is constrained (Jongman et al 1988). That was

the case here. The arch effect occurs for two fundamentally different reasons in CA

and CCA. In CCA diagrams it is usually caused by the inclusion of too many

environmental variables whereas in CA it a mathematical artifact caused by the

structure of the species by samples data set. Although detrending alters the position of

points in an ordination, their relative positions to one another should remain relatively

constant making it feasible to compare between ordinations.

River Etive

The length of the environmental gradient as determined by DCA was long (Axis 1 =

4.135 s.d ), so a unimodal based analysis was used i.e. CCA. The best four explanatory

variables were chosen using forward selection. The variables % bed rock (1 sample) ,

% fine gravel (6 samples) and % sand (1 sample) were not used in the analysis as all

occurred in very few samples. They were included as passive variables only. The taxon

Oxythira was used passively after demonstrating a much stronger relationship with axis

1 & 2 than all the other species.

70


Duneaton Water

The length of the environmental gradient as determined by DCA for the larger data set

was between 3 and 4 s.d. (Axis 1 =3.05 s.d.), the cut off point for selecting either a

linear or a unimodal based analysis. In this instance either a unimodal or linear method

is applicable, CCA was chosen. Log transformation of the species data reduced the

variance in the species data explained by the first two ordination axes to 11.4% and the

explained variance in the fitted species data from 78.7% and was not applied in the

final analysis. Extreme values were detected for 7 samples, but all ‘really belong to the

population’ (ter Braak & Smilauer 1998, pi 18) so these outliers remained in the

analysis. Of the possible variables used in the analysis, both % composition of boulder

and % composition o f fine gravel were excluded as both substrate types were found

very infrequently (n = 2 and n = 6 samples respectively) causing them to appear to be

collinear with the variable % composition of sand. None of the remaining variables had

a variance inflation factor >20 indicating that they all had a unique contribution to the

canonical ordination.

Blane

A DCA of the Blane Water species data set (40 useable samples) showed the gradient

length to be less than 3 s.d. (2.505) suggesting a linear as opposed to a unimodal

model was most applicable e.g. redundancy analysis (RDA). RDA analysis was not

significant for all canonical axes taken together (Monto Carlo test full model p =

0.085) although the first axis was significant (Monto Carlo test p = 0.015).

Constraining the ordination with environmental variables was not significant, so the

species data are presented in an indirect gradient analysis ordination, Principal

Components Analysis (PCA). Four samples had extreme values for some of the

71


environmental variables but did not reflect any detectable mistakes in data collection so

they were not removed from the analysis.

3.3 Results

3.3.1 Suitability o f data sets

For all sites, the data subsets containing variables derived from velocity profiles

produced ordinations with no significant canonical axes (Monto Carlo Permutation test

p > 0.05), see table 3.1.

Table 3-1, Eigenvalues and their significance from ordinations performed using parameters derived from velocity profiles as environmental variables. For the Duneaton Water and River Etive the ordinations were CCA, for the Blane Water the ordination was an RDA. Permutation tests were Monto Carlo tests carriedout using 199 randomly seeded runs.______________________________________River Axis 1 All

canonical

Axis

Eigenvalue F-ratio P-value Sum of all F-ratio P-value

eigenvalues

Blane 0.134 3.875 0.91 0.185 0.566 0.88

Water

Duneaton 0.141 1.951 0.665 0.539 0.997 0.435

Water

River Etive 0.133 2.003 0.33 0.45 1.056 0.345

72


3.3.2 River Etive

The DCA scatter plot (figure 3-1) shows two major groups as split by TWINSPAN on

its first iteration (TWINSPAN eigenvalue = 0.705). Further iterations produced splits

which had low eigenvalues and were not used (0.350 or less).

Monto Carlo tests of the first canonical axis o f the CCA and all the canonical axes

together were significant (p = 0.005, F-ratio 3.516 and 2.361 respectively). The biplot

(Figure 3-2) based on the first two canonical axes explains 11.3% of the variance in the

species data and 67 .4% of the variance in the fitted species data.

Sim

OulR ise

Pem i

Livo

Dia

Barh

Hep PebiOrth CanEcdyOlig

Lemo

DytTany

+0 . 0 Chir

0 . 5 + 5 . 5

Figure 3-1 Etive DCA. TWINSPAN group A is enclosed by the polygon, all other taxa are in Group B. See chapter 2 tables 2-6, 2-7 for a key to the species codes.

73


Axis 1 (eigenvalue 0.38) can be interpreted as a microhabitat gradient from erosional

to depositional conditions. Velocity and % cover of cobble are most strongly

correlated with Axis 1 ( weighted correlation 0.69 and 0.41 respectively). Depth and %

cover of boulders are both negatively correlated with Axis 1. Axis 2 (eigenvalue 0.24)

is most strongly correlated with % cover of boulder and cobble. The coarser material,

boulder, is unusually associated with slow flows, in reality boulders were found

throughout both sections but only in the deeper, slower areas, where they remained

submerged, and available to invertebrates. Only submerged boulders were sampled.

The smaller TWINSPAN group members are scattered throughout the right hand side

of the CCA diagram. Mayfly and stonefly species occur almost exclusively on the right

hand side of the CCA diagram as does the riffle beetle larvae Limnius volckmari. To

the right of the centre o f the diagram are Trichoptera larvae which require stable

substrates on which they can construct their filter nets. The left hand side of the CCA

plot is dominated by detritivorous Chironominae, Tanypodinae, Tipulidae and

Oligochaeta and the tiny Trichopteran genera Hydroptila and Oxyethira. These taxa

have known preferences for depositional areas, either for feeding or case building

material (Hynes 1972; Wallace et al. 1990). The presence of Orthocladiinae larvae in

samples with higher percentages o f cobble reflects a general trend within the

Chironomidae. Orthocladiinae prefer large substrate elements to the Chironominae and

Tanypodinae (Williams & Feltmate 1994). The Simuliidae were present at extremely

high velocities. Representatives o f both TWINSPAN groups occurred in each stretch,

with only two OTUs showed a strong preference for either deep or shallow stretches.

Scattered throughout the diagram are predators. The only cased caddis present were

members of the Hydroplilidae which are extremely small. Larger sized cased caddis

74


species were notable by their absence. Hence, the CCA ordination showed that the

observed differences in invertebrate occurrence could potentially be functionally

related to differences in habitat conditions which were paralleled in the literature.

o

B oulderCalu

PemiOuli

Rise • Cobble * .

/ SimuVelocity

Depth

Hydr I Epig- - - T

* • •Oxyt# Chir % |

Tipu Tany;

• BarhI-----Livo Can

• • Lemo Poki

Pebi

Hept

RhseOlig

Siar

1 . 0 + 1 . 0

Figure 3-2, CCA species-environment biplot for the R. Etive data. Variance added (A,k) for environmental variables: Velocity 0.32 (P = 0.005), Boulder 0.25(P = 0.005), Cobble 0.19(P = 0.015), Depth 0.15(P = 0.045), Gravel was not significant and was eliminated from the ordination. Only OTUs with an occurrence greater than 5 were used in the analysis. OTU codes are give in table 1-6 with the exception of the Chironomidae which were amalgamated in to subfamilies: Chironominae (Chir), Orthocladiinae (Orth), Tanypodinae (Tany) and Diamesinae (Diam).

75


3.3.3 Duneaton Water

The DCA scatter plot below (Figure 3-3) shows two major groups as split by

TWINSPAN (TWINSPAN eigenvalue = 0.784). Further splits by TWINSPAN had

low eigenvalues and were not used (0.345 or less).

For the final CCA, Monto Carlo tests of the first canonical axis and all the canonical

axis together were significant (p = 0.005, F-ratio = 7.373, 3.080 respectively). The

analysis used 5 environmental variables and orientated 28 species which occurred 353

times in the data set. The variance in the species data explained by the first two

ordination axes was 19.4% and the explained variance in the fitted species data from

was 67.7%.

76


Sim

RhdoLegeHyps

AgmuTip

BaseOliLemoChi

OutrBaerhoHyd

LehiHir Anne

LivoEpig

Gapu

Espa AnflOutu

Ecdy

Pemi

Sue

Glag

0 . 5 + 4 . 0

Figure 3-3, Duneaton Water DCA with TWINSPAN groups overlaid. Group A is enclosed in the polygon, group B includes all other taxa. See table 1-6 for OTU codes.

77

% Cobble

24.00 31.00

-.4995 .4109

.6931 .0223

1.0000%

Gravel

53.00 28.00

-.1308 -.5434

-.1309 .2465

-.5496 1.0000

% Sand

18.00 26.00

.7923 .0926

-.6234 -.2763

-.4968 -.2753

1.000

ocpT3 CP

PP<Ooo

~o £

o otOI—‘ to

a* oofl> rt < P— • 3& O-O3 Cl*

toov

toc/v-aVO

VOOvto

U)U )VO

oooo

to

CPpp<ooo

oCD

o x

noCT cr

ox00P3CL

& I 2 ©i *3 wP -*5T <!ceS* 2.' QTQ 3*<-K

CDa.n©*nCD

O3

3&2 .K‘T2.ire'sr

p3•ocTop

p2 .p 'O '

P9a.►

N»O

3“©nP3O3 'o’

o*nQ.S’pV*o’3C%O.

©c3C6P©3

acpT3

toVOVOcyi

-0Ov4L

OO

VOto

LO<01

OOOO

OvOOVOOV

to-PU)

toOv-oVO

oooo

o x

noCTcrCP

o

oVOVO

oU)VO00u>

oooo

oxwocatcp

o o I—* U)to to 4L U)00 u> o oo o

o o to 4Lto ■_ 4L -tLo vo o

o Oo

ou>toU)00

o O o o1—> to cyito Ov o U)OO - a to ■oVO to VO o

oooo

apP

CP ►—* •00O'r icpCL

00poCLp

CLCD<I ’oo

to

oxWocata

ox

OocrcrcT

CPpo<ooo

OCP

Table 3-2, W

eighted correlation

matrix

(weight

= sam

ple total)

for environm

ental variables and

Axis 1

and 2

of the

canonical ordination

used in

the R.

Etive

analysis.


As can be seen from Table 3-4, velocity is strongly positively correlated with cobble,

the larger more stable substrate element at the site, and is negatively correlated with

sand. Depth is not strongly negatively correlated with velocity as it often is at other

sites. Axis 1 (eigenvalue 0.404) is strongly positively correlated with the % sand in

samples and strongly negatively correlated with both mean velocity and % cobble.

Axis 2 (eigenvalue 0.175) is most strongly correlated with % gravel (negative) and

positively correlated with % cobble and mean velocity. The use of biplot scaling with

the focus on inter-species distances as used in the analysis of the Etive data resulted in

the species points clustering in the diagram centre making interpretation difficult and

the scaling was altered to inter-species scaling using Hill’s scaling (ter Braak &

Smilauer 1998). The species are grouped along axis 1 and to the left of axis 2 . The

Ephemeroptera and Plecoptera occur exclusively on the left hand side of the origin, the

other orders are scattered across the diagram depending on the preferences of their

composite genera.

TWINSPAN group A members, plotted in the Duneaton Water DCA remain

associated with one another in the CCA plot (Figure 3-4). Some group A members are

strongly associated with high percentages of sand and independent o f other variables

and can be found to the right of the diagram; Succinea, Hydrocarina, Chironomidae

and Ouliminius tuberculatus. Closer to the origin, and occurring in samples where

velocity and the percentage of cobble would be low, is Leptistoma hiratum and

Hirundinae. Closer to the origin where the mean of all variables occur are the

Oligochaeta and Gammarus pulex (which is a group B member). This suggests these

two taxa have a either a preference for the mean of all condition measured or are


Sim

VMo COBBLE

Outr*

LOCITY

Hydop

Suc SANDGlagLefJ Outu

BarhChi H?dLemo * japu

Oli *,LeGe G roup C y Lehi

Ecdy

LumAgmu Espa1

DEP™ Pemi.Ann<

GRAVEL

+4 . 01 . 5

Figure 3-4, Duneaton Water CCA. Group C enclosed by the box includes E. ignita, B. Scambus, L. volkmari and A. fluviatilis. Variance added by the environmental variables listed in order of their inclusion in the model; variance (A.A) values in brackets are the amount of additional variance explained by the inclusion of each variable sand (0.39 p = 0.005), cobble (0.14 p = 0.005), depth (0.09 p = 0.045), gravel (0.06 p = 0.205) and mean water column (mwc) velocity (0.06 p = 0.315).

ubiquitous in their presence. Anabolia nervosa, one of only two group A members not

closely associated with other group A members in the CCA shows a preference for

high % gravel and deep water. The other vagrant is Oulimnius troglodytes which has a

preference for high % cobble and fast water.

The taxa of group B show a more varied series of interactions with flow variables. At

the top left of axis 2 are two taxa, the Simuliidae and Rhyacophila dorsalis with a

strong affinity for high % of cobble and fast water. The Hydropsyche also have a

80


strong preference for fast water but their preference for high % of cobble is less. These

3 taxa can be found close together in the DCA plot. To the left of the origin there is a

group with no strong affinity for either velocity or depth. This grouping does prefer

samples with less sand and more of the other substrate types, it includes;

Glossomatidae (possibly Agapelus), Leuctra sp., Baetis rhodani, Baetis scam bus,

Ephemerella ignita, Ancylus fluviatilis and the Tipulidae. It is the largest of all the

groups. Strongly associated with deep water are the Lumbricidae, Agraylea

multipunctata and Esolus parallelepipedus.

3.3.4 Blane Water

The data used in the PCA covers 19 taxa which occurred, cumulatively 254 times in 40

samples. The variance inflation factors for all environmental variables were low (all <

2) suggesting that the variables are not highly correlated with one another which

agrees with their product moment correlations in Table 3-4, all o f which were less

than 0.5. Depth is positively correlated with both velocity and high percentages of

cobble. Velocity is not strongly positively correlated with % cobble as was expected

although it is relatively strongly and negatively correlated with % sand.

81

oha

SO so vOO ' O ' O ' O 'cdp3

<2*Oo

C/3 T lg 5 g. «OQ•~i

P<CD

nocrSicT

o o 1—1 On On •o1—1 4L to 4L LA ONJL to LA to 00 oto 00 -0 00 NO U>►—1 00 U)

toJL JL

LA 002 .

o q "5 “CDCL

3cdp3

HP2«r<*>

2srBftfta

oBsCDB

«!P3.695T

o o t—1 b—I to NOb U) O n o -o ON00 to 4L h—‘ JL i—»NO o NO to 4L oON o o to NO o-o to - J

Ot ot oLAU)

o O Ob b toU) NO 1— 1LA 4L Lk)00 00 -tL

o o o oto ’o b toto to u>t—1 00 NO LAo u> LA O

: O8\ OJ

o o •-*b P* oU> no o4L O O—] On O

o o o oH-* O~o o

oooo

c/3r +p3CL

a*CD<Bo3

ox

noc rct;cT

ef'

xoxrrj5’CD

OQ>-t%CD

32.il*<■*■ft

c .

BfttoB

PBa.pQ.a.CD■<

OBPBa.

T9*noa.B

oo 3 8"-J o Oi o

o xC/3P3C L

\ P3 tO8 NO oooo

i CD i P! 3I <f 2 * I o S 2.( r-KI'C

5 O ; O 8 O 8 O

! O! CD T3er

Com

munity

ordination


Glbo

AgfuBarh

MnVelRhdo

Anfl Hysi

Gapu

Li vo Elae Asaq

PoflDepth

Fine graOulPspu

Poje Epda

DyOr

Sand

- 1 . 01 . 0 +1.0

Figure 3-5, Blane PCA correlation biplot with environmental variables imposed over the ordination. Solid arrows = environmental variables, dashed lines = taxa. Scaling focuses on intcrspccies distances and the ordination diagram displays standardised species data (by dividing by their standard deviation) post transformation, and correlations. The values that can be inferred by the biplot rules are thus correlation coefficients(ter Braak 1997,1998)

The biplot in Figure 3-5 can be interpreted using the rules given by Jongman et al

(1987: p i27). The arrow for each taxon indicates a vector increasing in the direction of

increasing abundance for each species, arrows pointing in the same direction are

positively correlated, perpendicular arrows reflect a lack of correlation between the

variables represented by the vectors.

The species are scattered throughout the diagram with little grouping apparent. The

distribution of Limnius volckmari and Elmis aenea are closely correlated as are

Ancylus fluviatilis and Agapetus fuscipes. Glossoma boltoni increases in the same

83


direction as gravel, as does Rhyacophila dorsalis with mean velocity. A group of 3

species with similar distributions are Baetis rhodani, Hydropsyche siltalai and

Gammarus pulex. Potamopyrgus jenkinsi and Ephemerella danica are also similar in

distribution.

3.4 DiscussionThe constrained ordinations shows that clustering of the invertebrate data was limited

to two major groups at the Duneaton water and the R. Etive sites. The flow variables

in these constrained ordinations showed strong relationships between some taxa and

flow variables. The often observed selective occurrence of macroinvertebrates at either

erosional or depositional conditions were also observed here although, as found by

Barmuta (1989) community structure appeared more continuous in nature and

confirms the view of earlier researchers that the use o f substrate as a basis for

delimitation bottom fauna may be limited (Thorup 1966). These selective occurrences

reflected the preferences of most taxa for slow or fast reaches as observed in the

previous chapter.

The implications for IFIM based studies are profound. Although some taxa showed

similar responses to flow variables in the different rivers e.g. Rhyacophila dorsalis

others were inconsistent. This reinforces the argument that data collected at one site

should not be used to build a model for other locations. It also highlights that there

can be considerable variation in the response of a taxon to flow variables.

Improvements to sampling strategy for future work in this area are suggested and the

implications for how we should view the patchy nature of the stream bed are discussed.

Clustering of the taxa data from the Duneaton Water and R. Etive with TWINSPAN

split both site communities into two main groups. For each site the composition of the

two groups produced was very different, probably reflecting site specific factors. The

84


smaller Duneaton Water group preferred deposition conditions whereas for the R.

Etive the smaller group was erosional in nature. One possible criticism of these results

is that as both groups contained a mixture of functional feeding groups (Group A from

the R. Etive contains Simulliidae and Baetis rhodani, respectively a sedentary filter

feeder and a swimming deposit feeder (Williams & Feltmate 1994)), it would be

expected that such groups would be separated by a cluster analysis on data with

sufficient habit resolution.

The limited success of TWINSPAN at splitting the communities into a greater number

of intelligible groups could be for one of two main reasons, either the communities are

relatively homogenous in composition and spatial distribution or alternatively

individual quadrats covered a sufficient range of underlying microhabits that

heterogeneity was disguised. Either way this is good news for those using quadrat

sampling as a representative measure of invertebrate community structure at a

particular site. In future work a sampling program based on smaller quadrat sizes

would be expected to yield clearer results as it has been frequently shown that the

distributions of benthic invertebrates are aggregated (Elliot 1977).

The data from the Blane Water proved more difficult to analyse and interpret probably

because of the more homogenous nature of the sampling points in this site. The range

of depths was less than at the other sites and the substrate was always mainly gravel,

see Chapter 2, Figures 2-7, 2-10. As the DCA examination showed the taxa appeared

to follow a linear rather than a unimodal response to the environmental variables

measured supporting the reasoning above.

Constraining ordinations with flow variables illuminated the individual taxon’s

preferences. Although the amount o f variance explained by the direct gradient

ordinations (CCAs) was low, suggested that the variables used are important

85


influences on the spatial distribution of benthic invertebrates in Scottish rivers and

elsewhere. A gradient between erosional and depositional conditions was displayed in

both constrained ordinations along which some of the taxa exhibited preferences

consistent with previously published data.

The small amount of variance in the species data explained by the ordination diagrams

suggests that the variables selected are only one of many factors influencing the

distribution of benthic invertebrates in the data set. That the multivariate environment

of stream benthos is complex has frequently been recognised and its potential for

limiting the success of such studies noted e.g. Williams & Smith (1996) endorses the

sentiments of (Hart 1992):

“that meaningful models o f lotic community structure will only arise if researchers

acknowledge the multifactorial organisation and dynamic nature o f these

communities ”.

It has been pointed out previously that species abundance data are frequently very

noisy leading to a small percentage of the variance in the data being explained (ter

Braak & Smilauer 1998: p i21). But it has also been stated that an ordination diagram

that explains a low percentage of the data can be quite informative (Gauch 1982) e.g.

the use of CA on benthic invertebrates data with comparable eigenvalues to the data

presented here showed clear patterns in distribution in Italian river systems (Rossaro &

Pietrangelo 1993).

There is another possible reason for the low amount of variance explained by the

environmental variables, that is they were not measured on a physical or temporal scale

suitable to easily detect invertebrate preferences. Downes has repeatedly shown that

finer scale measurements o f environmental variables yields illuminating results

(Downes & Jordan 1993; Downes et al. 1995) as have others (Whetmore et al. 1990).


Unfortunately these papers have concentrated their fine scale measurements on larger

substrate elements to the exclusion of other substrate types (sand, fine gravel) and

therefore were not applicable here. Others have suggested that invertebrate distribution

is a function of flow events that have preceded the current conditions (Clausen &

Biggs 1997) or that conditions directly upstream can affect spatial distribution (Quinn

et al. 1996). Although this is a persuasive argument, when one thinks o f the

disturbance caused by flooding it is not wholly convincing particularly in light of the

rapid ability of invertebrates to recolonise the stream bed post disturbance (Brooks &

Boulton 1991). This rapid response is somewhat dependent on the slope and bed

structure o f the river (Armitage & Gunn 1996). A final possibility is that different

instars or species within a taxon had different habitat preferences but when

amalgamated together under their taxon name these habitat preferences were hidden. It

has been shown that small nymphs and large nymphs o f Ecdyonurus species, a taxa

occurring in this study exhibit preferences for different bed roughness (Buffagni et al.

1995).

There are many other possible reasons for the small amount of variation explained but

they are not related to flow variables, e.g. predation or periphyton availability

(Dudgeon & Chan 1992; Lancaster et al. 1990). It is not possible to discuss these

factors in detail because no measurements of such interactions are recorded. Finally the

possibility exists that the animals are not strongly influenced by the environmental

variables measured.

That no additional variation in the data was explained by variables derived from

velocity profiles was disappointing. Although others have successfully incorporated

such measures into multivariate analysis of similar data (Quinn & Hickey 1994) I

conclude that estimates of near-bed flow conditions derived from one velocity profile,

87


measured over one spot on a river bed in a standard 25 x 25 cm quadrat are not

applicable to the whole of that quadrat. The complex range of near-bed flow patterns

that can occur in one quadrat would be the cause. The more traditional use of mean

water column velocity and depth are more likely to be indicative of the incident flow

arriving over a qaudrat and is applicable to the whole quadrat because it is not hugely

effected by the bed form directly below. Measures of substrate type are important to

complete the picture. The advent o f new technology (acoustic doppler velocimeter) has

allowed the rapid measurement of turbulence in 3 dimensions above the stream bed and

has been shown as an important variable in the distribution of invertebrates (Bouckaert

& Davis 1998). Although heeding the advice of Pringle et al. (1988): advise cited in

the introduction (chapter 1) one imagines the incorporation of measures o f turbulence

into future sampling plans would be highly illuminating. Interestingly, test ordinations

of the velocity profile variables showed the R2 values for the profiles explained

additional variation in the distribution of invertebrates (data not shown). As the R2

value can be interpreted as a measure o f the deviation o f the flow structure above the

bed from the a standard log-normal relationship it can be seen as a measure of gross

turbulence (which would not be reflected in Re) and may provide a useful additional

measure of habitat complexity.

3.5 Conclusions• Unconstrained ordinations did not produce clear aggregations of taxa, mirroring

some other studies.

• Constrained ordinations did show a gradient of erosional to depositional conditions

with some taxa showing preferences for these conditions.

• Responses of some taxa varied between rivers as did the relationships between

physical parameters measured, implying that models produced for one river are

88


likely to be site specific, limiting the use of IFIM based PHABSIM studies in river

management o f invertebrate habitat

• A re-evaluation of the importance of flow parameters is necessary with the

inclusion of turbulence measurements, spatial scales tailored to individual species

and more emphasis on alternative effects o f flow parameters, e.g. spatial

redistribution of invertebrates, conditions for periphyton growth etc.

89

Individual responses

Chapter 4: Individual responses to hydraulic parameters

4.1 Introduction

This chapter concentrates on the responses of individual taxa to the environmental

variables measured in Chapter 2. The multivariate techniques used in Chapter 3 are

based on the assumption that species respond to environmental gradients either in a

linear or a unimodal manner. As the methods used are robust enough to deal with

noisy data (Jongman et al. 1988) few users actually publish the underlying response

curves of the species o f interest. This pragmatic approach has left significant gaps in

our knowledge. As mentioned in the previous chapter, management tools such as

PHABSIM make a number of assumptions about the responses o f animals, including a

constant or fixed response curve to flow (Milhous et al. 1984).

Finding information to support these assumptions is very difficult simply because there

is such a small amount of published data on species responses to individual flow

variables. Only three references were found (Jowett et al 1991; Lancaster1999;

Fjellheim 1996) which give response plots of invertebrates to flow variables in the field

using measures of abundance. Based on presence absence data others have provided

logistic (or logit) regression analysis curves, plotting the probability of gammarid

species occurring along various environmental gradients (Peeters & Gardeniers 1998)

including water velocity. The technique has also been successfully applied to the

distribution of water beetles, although the interest here, was mainly in lentic species

(Eyre et al. 1992; Eyre et al. 1993).

There have been a large number of papers examining the flow preferences of

freshwater benthic invertebrates in the field. Some older studies do look at individual

variables but the results tend to be descriptive only (Jones et al. 1977). Of the more

recent works few look at variables singly instead, analysis tends to be multivariate

90


(Downes et al. 1995; Hawkins et al. 1997; Puckridge et al. 1998). This makes sense

when one is dealing with a large number of variables and species, especially as many of

the variables may have synergistic effects. However examination of responses of

individual species or taxa allow us to ask some important questions.

The first question is whether or not the animals actually respond to the variables at all

and if they do, do they conform to our views of niche occupation. In attempts to show

taxon preferences, abundance can be plotted against an environmental variable and a

curve fitted.

If the environmental variable data was collected using a randomised sampling regime

any curve fitted to the data is highly dependent on the distribution of the

environmental variable. For example a taxon may prefer a velocity of 0.5 ms"1 but 0.2

ms"1 is the most common velocity sampled and lies within the taxon’s velocity

tolerance range. Then the taxon’s maximum abundance could occur at 0.2 ms'1. A

fitted curve describing this type of response represents the taxon's compromise

distribution.

Data collected in this manner has been used to produce in-stream flow-habitat

suitability curves for benthic invertebrates (Jowett et al. 1991). So the possibility exists

that the invertebrate’s actual preferences are not being identified and the models are

based on incorrect assumptions. This is more likely to happen if the species have a

broad tolerance for the environmental variables in question and also covers the

available range of the environmental variable. Unfortunately these studies do not

report the distribution of the environmental variables of interest so it is impossible to

judge the quality of their results.

Here abundance of individual taxa is plotted against depth, velocity and substrate type

with frequency histograms of these environmental variables also present. The

91


relationship between the environmental variables and taxon abundance was tested in

two ways for each taxon.

Where possible the following hypotheses were tested:

H'0: That the animals were responding to individual flow variables in a gaussian

manner, this response includes both linear and unimiodal response types.

H20:That the observed responses were not artifacts of the sampling program and did

indicate an actual response by the invertebrates.

It was not possible to test the second hypothesis conclusively. But it was believed that

if the maximum abundance of a taxon in individual samples (dependent variable) was

correlated to the uneven sampling of the environmental (independent) variable that this

might indicate any relationship was a sampling artifact.

By this I mean that variables such as velocity, if recorded at random sampling points in

a stream will show a strongly schewed distribution, e.g. some velocities are being

recorded more frequently, their ‘abundance’ is higher. If velocity does not have a large

influence on the distribution of a particular species it may still show a positive

correlation with the more common velocities simply because these represent more

frequently sampled points on the river bed and with each sample ones chances of

recording the animals at high densities increases. Therefore I predict that a given taxon

will reach its maximum abundance in samples in which the environmental variable also

reaches its maximum abundance e.g. if the most frequent velocity was 0.25 ms'1 then

the species of interest would reach its maximum abundance in areas of the river where

that velocity occurred.

To ecologists there is an obvious alternative reason why a species would occur at the

most frequently available point along an environmental gradient and that is because it

is highly adapted to it! However some of the variables measured, such as depth and

92


velocity alter on a daily or even hourly basis in rivers so it would be odd that the

animals were exactly adapted to the conditions on the day, but not on the previous day

for example. It was not possible to exclude this possibility from the analysis but it is

hoped that by presenting the responses o f the animals at each site separately,

comparisons of responses at the different rivers would aid the evaluation o f this

possibility.

4.2 Methods

4.2.1 Data

The data collection methods are given in Section 2.2 of Chapter 2. Not all taxa

sampled were used in model fitting. Those excluded had low occurrences. The samples

taken were quantitative and are represented as numbers of individuals per sample in all

cases. Actual number of animals recorded is used in the response plots, so the reader

can clearly see the number o f individuals each plot is based upon.

4.2.2 Statistical Analysis

To test the two hypotheses each environmental variable was divided into a number of

intervals and the frequency of each interval calculated (frequency = no. of samples in

which that interval occurred). The intervals for each variable, at the three rivers are

given in tables 4-1, 4-2 and 4-3. Intervals were chosen using a pragmatic approach

which attempted to assign the same number of intervals to the same variable at all

sites, which was not always possible due to the distribution o f the variables. It is

possible that by increasing the number of intervals I increased n and the possibility of

getting a significant correlation but if I had given the same number of intervals to all

variables their occurrence would have been too low in all categories to show any

responses. The maximum sample abundance of each taxon for each interval was

identified using an excel pivot table. By maximum sample abundance for a particular


interval I mean the density of individuals occurring within a sample which has the

highest density o f individuals for all samples whose environmental variable measures

fall within that interval. Least squares regression was performed with the frequency of

a given variable's intervals as the independent variable and the maximum density of

individuals per interval as the response variable. This analysis was repeated for each

taxon at each river. The results of this analysis are presented in a general review of

taxon responses to environmental variables which also includes non - significant

responses. The totalled abundances were also examined to test the second

hypothesis and represent the summed abundance for a particular interval.

94


Table 4-1, Velocity (mean water column velocity) intervals used in regression

analysis.

R. Etive

Intervals

ms' 1Frequency

DuneatonWater

Intervalsms' 1

Frequency

Blane Water

Intervalsms

Frequency

0-0.05 7 0-0.05 14 0-0.1 60.05-0.1 8 0.05-0.1 4 0.1 - 0.2 40.1-0.15 7 0.1-0.15 6 0 .2-0 .3 40.15-0.2 6 0.15-0.2 5 0.3 -0.4 100.2 - 0.25 2 0.2 - 0.25 6 0.4-0.5 20.25-0.3 5 0.25 - 0.3 1

oo'io

40 .3-0.35 2 0.3-0.35 6 0.6 -0.7 40.35-0 .4 2 0.35-0.4 1

0©o'1r-o' 10.4 - 0.45 3 0.4 - 0.45 2 0.8-0 .9 10.55-0.6 2 0.5-0.55 1 0.9-1.0 20.6-0.65 2 0.55-0.6 1 > 1.0 20.65-0.7 3 0.6-0.65 2 X

>0.7 2 0.65-0.7 2 X


Table 4-2, Depth intervals used in regression analysis.

R. Etive Duneaton Water Blane WaterInterval Frequency Interval Frequency Interval Frequency

m m m0-0.05 3 0-0.05 6 0-0.05 60.05-0.1 3 0.05-0.1 9 0.05-0.1 110.1-0.15 6 0.1-0.15 8 0.1-0.15 60.15-0.2 10 0.15-0.2 7 0.15-0.2 70.2 - 0.25 7 0.2 - 0.25 9 0.2-0.25 60.25-0.3 3 0.25 - 0.3 4 0.25-0.3 10.3-0.35 4 0.3-0.35 3 0.3-0.35 20.35-0.4 4 0.35-0.4 3 >0.5 10.4 - 0.45 1 0.4 - 0.45 30.5-0.55 4 0.45-0.5 10.55-0.6 10.6-0.65 10.65-0.7 1>0.7 3

96


Table 4-3, Substrate intervals used in regression analysis. Percentage composition of standard substrate units, cobble, gravel etc., were transformed to average length measures in mm by taking the mean length of each substrate category and using it to get the weighted average for each sample._______________

R. Etive DuneatonWater

BlaneWater

Interval Frequency Interval Frequency Interval Frequencymm mm mm

0-50 11 0-20 12 0-10 150-100 11 20-40 13 10-20 2100-150 4 40-60 6 20-30 7100-200 5 60-80 5 30-40 12200-250 7 80-100 9 40-50 10250-300 12 100-120 1 50-60 1

>120 7 >60 5

I attempted to fit Gaussian curves to the abundance data directly using a version of the

method suggested by (Jongman et al. 1987): c.f. Section 3.4 ‘Regression for

abundance data with many zero values'. They advocate the use of log-linear regression

where the Gaussian curve takes the form:

log(y) = b0 + bjx -+ b2 x2

where b2 <0. As the term b0 + bix + b2 x2 is a linear predictor it allows the use of

multiple regression applications in statistical packages where x and x2 can be entered

as separate variables (Lancaster, J. Edinburgh University pers comm ). This was the

method used here. There is an alternative form of the equation which has some

attraction for biologists this is:

log(y) = a -0.5(x - u)2 / t 2

such that ii is the optimum position of the species along the environmental gradient

(value of x where the maximum abundance occurs), t is its tolerance (a measure of

ecological amplitude) and a is the log of the species maximum abundance (c) (ter

Braak & Prentice 1988).

97


It is possible to estimate these parameters using the following conversions:

the optimum, u = - b,/(2b2)

the tolerance, t = 1 / ^(~2b2)

the maximum, c = exp(b0 + bjx + b2 x2)

4.3 Results

The results are divided into three main sections; one for each environmental variable.

These sections are subdivided by river and at the end of each section, comments on

taxa which occur in more than a single river are made. Presentation o f information as

basic scatter plots of individual taxa against environmental variables aids interpretation

of the statistical analyses. An introductory review of these analyses is followed by the

presentation of the results by variable and by taxon.

4.3.1 Curve Fitting

Attempts to fit Gaussian curves to the data produced mainly non-significant results, in

spite o f the apparent unimodal responses suggested by many of the scatter plots o f

abundance against environmental gradients, see Appendix V. Of the cases where

curves fitted were significant see Figure 4-1.

4.3.2 Relationships o f totalled maximum abundance to the frequency o f variable

intervals

For the majority of taxa which did show a significant response, the correlations were

moderate; for 26 out of the 31 significant correlation’s, the R 2a<ij value was less than

0.75, figs 4-2a, 4-2b, and 4-3. Of the 3 variables examined, substrate had the fewest

significant correlations (6), all o f which were for the Blane Water. Depth intervals

were most frequently correlated (14) and velocity intervals were the second most

98

Log

abun

danc

e Lo

g ab

unda

nce

Log

abun

danc

e


River Etive y = -0 .53 + 5.08*x - 6.53*x*x

0.5

0.0

D epth m T anypodinae

D uneaton W ate r y = 1.42 - 8.55*x + 13.79*x*x

-0.1 0.1 0.3 0.5 0.70 .0 0.2 0 .4 0.6 0.8

Velocity m s"1 Oulimnius tuberculatus

River Etive y =0.83- 3.75*x + 3.69*x*x

0.8

0.4

0.0

■0.1 0.1 0.3 0.5 0.7 0 .9 1.1D epth m

Polycentropus flavomaculatus

D uneaton W ate r y = 2 .6 2 -8 .9 4 * x + 12.16*x*x

4.5• _ •

3.5

2 .51.5 #0.5

-0 5

• • • • « • • • •

-0.1 0.1 0 .3 0.5 0.70 .0 0.2 0 .4 0 .6 0.8

Velocity m s"1 C hironom idae

Blane W ate r y = -0 .33 + 3.37*x + 0.0004*x*x

3

2

1

01 ..........................................0.05 0.05 0.15 0 .25 0 .35

0.00 0.10 0 .20 0 .30 0.40D epth m

Ancylus fluviatilis

D uneaton W ate r y = 1.03 + 8.87*x - 13.23*x*x

5 4

0) oo 3 c(0 n

T3 Z C

j> 1 (0D) 0 O- 1 -1

-0.1 0.1 0.3 0 .5 0.70.0 0.2 0.4 0.6 0.8

Velocity m s '1 Ecdyonurus spp

Figure 4-1. Plots of log abundance against depth or velocity for animals exhibiting statistically significant gaussian relationships.

99

Figure 4-2a,

Regression

scatter plots

of m

aximum

sam

ple abundance

against frequency

of velocity

intervals. Ephemerella

ignita, p

0.01, R2adj =

0.47 F

= 11.73, n

= 13; H

ydroptila spp., p

< 0.01, R

2adj = 0.58, F

= 18.10, n

= 13;

Chironom

idae, p <

0.001, R2adj =

0.69, F 28.9, n

= 14

and O

ligochaeta spp., p

< 0.01, R

2adj = 0.54, F

= 15.41, n

= 14.

Maximum ab u n d a n ce values

3X)cd>3O< is> 5 r* a> oi a1 1 3 o a 2.oi ^

Maximum ab u n d an ce values io u A 01

Maximum ab u n d a n ce values ) —» oj a 01 o

Maximum ab u n d a n ce values

N o w t u a i o M t O )

« ° « co 3

a> s- * ® X ->

: <l> Oi

3C l

D.303

R. Etive

R. E

tive -.37+

0.58*x y

= -2.25

+ 1

,75*x

Figure 4-2b, R

egression scatter

plots of m

aximum

sam

ple abundance

against frequency

of depth

intervals. Orthocladiinae

spp. p <

0.01, R

2«dj = 0.44

F =11.38, n

= 14; Baetis

rhodani p <

0.05, R2adj =

0.286 F

= 4.8, n

= 14; Baetis

spp. p

< 0.05, R

2adj = 0.45, F

= 8.46, n

= 10;

Limnius

volckmari p

< 0.001, R

2„dj = 0.81, F

= 40.31, n

= 10;

Oulimnius troglodytes

p <

0.05, R2adj =

0.39, F =

6.85, n =

10; Tipulidae. p

< 0.05, R

2adj = 0.34, F

=5.67, n =

10 and

Oulimnius

spp. p <

0.05, R2adj =

0.54, F =

9.31, n =

8.

Mawtum abundance values Maximum abundance values Maximum abundance values

Maamum abundance values

abundance value?


Figure 4-3, R

egression scatter

plots of

maxim

um

sample

abundance against

frequency of

substrate intervals. H

ydroptila spp.

p <

0.05, R

2adj = 0.68, F

=12.04, n =

6; Chironom

idae spp.

p <

0.05, R2adj =

0.63, F =11.55, n

=7;OH

gochaeta spp. p

< 0.05, R

2adj = 0.70, F

=15.40, n =7;

Tipulidae spp. p

< 0.0001, R

2adj = 0.97, F

=201.0, n =

7; Leuctra

spp. p <

0.01, R2adj =

0.85 F

= 35.77, n

= 7; Lim

nius volckm

ari p

< 0.05, R

2adj = 0.57

F =

8.82, n =

7; Oulimnius spp.

p <

0.05, R2adJ =

0.54, F =

8.18, n =

7.

Maximum abundance values Maximum abundance values



frequently correlated (11). The R. Etive had the fewest significant conditions,

reflecting the limited number o f taxa which occurred in sufficient abundance to be

assessed. The Duneaton Water had fewer significant relationships than the Blane

Water.

4.3.3 Relationships between maximum sample abundance to the frequency of

variable intervals

Significant relationships were limited to a small number of the taxa. With the exception

of substrate, which had 7 significant correlations, the number of correlations was fewer

than for the relationships between totalled abundance and the environmental variables

depth (7) and velocity (4): Table 4-4.

There were no significant correlations for velocity at the Blane Water which was

represented among the sites with significant relationships for the other two variables. It

had the fewest significant relationships o f all sites with 4 in total while the R. Etive had

5 and the Duneaton Water had 9.

103

Duneaton W

ater Chironom

idae <0.05

(1,5) 11.14

0.62 7

-23.46 11.47

Ephemerella

ignita <0.01

(1,9) 15.22 0.58

11 24.75

0.78O

ligochaeta <0.05

(1,5) 9.15 0.57

7 -0.37

2.93Oulim

nius tuberculatus

<0.05 (1,9) 8.38

0.42 11

30.86 .

72

ocdXS

m<'cd

Wp *s

Pn>

T3Cpa>

O3croop*Bt3 ’pcd

toft 204&3-0

1

a*•>4.ataa<>3

a*

Ci>!SS&

s

^ sCl* 3 PCD aCn

C§

(S'Oj

OQOocrpcdpfp

acdCD&o3

Bcd

O3t►-toaogalpa>

• C>c nS £*5: ^ pCD

<SLoo

Wp-+-

CD

>Ta S'S’ cTa

a*C&

3C>

&f5"

A A A A A A A A A A A A A A AO O O 0 O 0 O O O 0 O O O O OO O O 0 ’0 0 O O 0 ’0 O O O O OO N O 1—>* t— * 1—> O n o n O n O O O O ►—* O 1—1

1—* 0 1 -^ O1—1 >— >

1— 1 1— > t— >

1— * 1— * VO VO VO VO f—i i—* _ ito to to 1—1 I—* 1—* •—> 1— * i_ i ►—* *—*

'— ■/ '— ' 1—1 O n 00 ^ ^ s— ■' '— - '— ' '— - '■—' ' w 'o n l—l VO O to b> rr\ O n to VO VO VO 1—* O n 1—1l—l O VO 41* '4 •—* 44* •—» 0 VO to 1 4^ O n ►-*1—* ON VO U ) 1— * 0 1 O OO b v 0 4^ VO 4 .

t—» VO OO O n to 00 to

0to 0 0 0 0

Lo 0 0O n O 0 0 0 0 0

to&

to00 On00044* 400 OnVO 4-0 4CO

0u> 1—14 00

VO<1co ■01—1 Onto OOto 4

O n

31—14 l—l 4 1—1 »—i1—1 1—1 l—lOJ CO 1—1co l—lOJ CO CO l—lCO 1—1VO bo

0.36-2.84-0.93-2.93-1.213.27

18.095.5

10000

-1.05-52.25

-1.22-0.706

-6.76-0.8

to to0 l— l 0 t o V O l— l ,— i H-1 O N O O n

1— 1

o> 0

to - 0 NO O n t o O n 0 to 1— 1 NO O n 0 0 VO- 0 - 0 O n to 4 s* 4^ VO ON ON 45* 00 1— 1 to ■ o

<P

a*

•nonr+Pr—t-OO ’o’

xto3

Table 4-4, The

results of the regression

analysis of totalled taxa

abundances (dependent variable) against frequency

of environmental gradient intervals

(independent).

C/0ccr

Sp*S

SCD

sp*g

R<D

oSi33S'Oj

; Ja : 3

S- S 3 § s S s- 3 g* D PT«ri?►£s3

a-ns

.*■*. aSi*N4

"5s

0)'•s.*»l.<A2-o

Sic>I

sa*»•*.tn<O

a-§si"HSiOj

xscalpa

A A A A A A A A A A A Ao o o o o o o O O o o O'o o ‘o o o o o o o o OoLA

L n oLA

oLA 1—1 o

LALA LA LA o

LAoLA LA

J —* I—* •—* >—1 i—i l—l l—l l—l l—l i—i i—i l—l"l a LA "l a "l a "l a ON ON "o n "o n ON ÔN "l as —^ s N—S s—^ '■ws s—^ sLk) i—i Lk) to to i-^ i 0 0 -o O n i—i i—i <3-o i—i Lk) LA o NO to ON 4 ^

OO |—1 i—i'4^ LA ON o 0 0 LA - O - J to LAto 'O LA •—1 - P to Lk) OO

O00LA £

90

0.8

4

0.8

0.7

6

0.7

2

0.5

1

0.4

8

0.4

3 i—iON

LANOP

0.5

7

- o ~ o ~ o ~ o OO OO OO 00 00 00 - o

1l-^l

1Lk)

■00

■ ■t o

1t o t o

1l—l P -2

.6

io

t oi—i

OOPOO

O nPO n

~ ot o ~ o

t oLk)

<1 i Lk)

t o Lk) O n Lk) t o t o u > Lk)t oLA t o Lk) t o

- Pt o

Oi—i

O

t oo00

O ni—i

NOON

op

i—i►—i

oNO

- OLA

NOLk)

cro

cr 3

Variable River

Taxon p

value F

statistic


1413

</)coOJEa> </>

-Q OO

I Ioz

oII >-'tN<*)'finiDPaico

7

65

4

3

210■1-0 1 0 1 0 3 0 5 0 7 0 9

7

65

4

3

2

1

0•1-01 0.1 0.3 0.5 0.7 0.9

V eloc ity m s r1Velocity m s-1 B ae tis rhodan i

Velocity m s 1 C hironom inae

5 5

4.5

oiu

X IEDZ

-0 1 0 1 0 3 0.5 0 7 0 9

12

108

6

4

2

02•0.1 0 1 0.3 0.5 0 7 0 9

22

w(0u"O>TJCotf)oz

-01 0.1 0 3 0 5 0 7 0 9

Velocity m s 1 E phem ere lla ign ita

Velocity m s 1 HydroptHa

Velocity m s~1 O ligochaeta

12

10

8

6

4

2

0-2-0 1 0 1 0 3 0 5 0 7 0 9

3 5

O

0.0

-0 1 0 1 0 3 0 5 0 7 0 9

3 5

3 0

2 5

2 01.5

1 00 5

0 0-0 5

-0 1 0 1 0 3 0 5 0 7 0 9

Velocity m s- 1 O rthoclad iinae

Velocity m s 1 Tanypodinae

Velocity ms~ T ipu lidae

Figure 4-4, Etive velocity scatter plots with a histogram showing the frequency distribution ofvelocity at sampling points. Abundance values are number of individuals per sample.

106


20

18

m 14CB 12 <0I 10 </>o 8oi/>oz

o cn n ■« m (o P s

j/t TO 3■g>"Oc'oui

76

5

01-0 1 0 1 0.3 0 5 0 7

Velocity ms

0 0 0 2 0 4 0 6 O.i

Velocity m s*1 Ancylus fluviatilis

14

10nI 8 >T3 6coinOz

-0 1 0 1 0 3 0 5 0 70 0 0 2 0 4 0 6 Ol

Velocity m s-1 Baetis sp.

65

55

45_C/>

3 35 >Y 25

omoz

■0.1 0.1 0 3 0 5 0 70 0 0 2 0 4 0 6 0 8

Velocity m s-1 Chironom idae spp.

70

60

50

| 40

■§ 30TOc— 20o

-10-0 1 0 1 0 3 0 5 0 7

0 0 0.2 0 4 0 6 0.I

Velocity m s-1 E cdyonurus spp

35

jn™ 25■g>TOE 15o

-0 1 01 0 3 0 5 0 70 0 0 2 0 4 0 6 0 i

Velocity m s-1 Ephemerella ignita

2 5

o

8 05 z 0 0-0 5

-0 1 0 1 0.3 0 5 0 70 0 0 2 0 4 0 6 01

Velocity m s’1 H ydropsyche sp.

28

22

COoj 16 ■g1 10ooz

-0 1 0 1 0 3 0 5 0 70 0 0 2 0 4 0 6 0 1

Velocity m s"1 Leuctra spp.

8

7

6

5

4

3

2

1

01-0 1 0.1 0 3 0 5 0 7

0 0 0 2 0 4 0 6 O.i

Velocity m s- 1 Limnius volckmari

F ig u re 4-5a , D un ea to n W a t e r ve loci ty s c a t t e r p lo ts w ith a h is to g ra m sh ow in g th e f r e q u e n c yd is t r ib u t io n o f velocity a t s a m p l in g poin ts . A b u n d a n c e values a r e n u m b e r o f in d iv id ua ls p e rsam ple .

107


20

oii

U)cuD■g>TDC

if)o

4.5

3.5

2.5

1.5

0.5

-0.5

OO o

o o

O O o

(OOOOOOO O O o o o

V elocity m s-1

-0.1 0.1 0.3 0.5 0.70.0 0.2 0.4 0.6 0.8

V eloc ity ms1 O ligochaeta sp.

8

7

6

5

4

3

2

1

0

10.1 0.1 0.3 0.5 0.7

0.0 0.2 0.4 0.6 0.8

V eloc ity ms1 Oulimnius tubercu la tus

<j)cuD■g>TDC

if)o

18

14

10o o

6

o o CD

2 OO

•o

20.1 0.1 0.3 0.5 0.7

0.0 0.2 0.4 0.6 0.8

V eloc ity ms Tipulidae spp .

F ig u re 4 -5b , D u n ea to n W a t e r ve loci ty s c a t t e r p lo ts w ith a h is to g ra m sho w ing th e f r e q u e n c yd is t r ib u t io n o f velocity a t s a m p l in g points . A b u n d a n c e v a lues a r e n u m b e r o f in d iv id ua ls p e rsam p le .

108

3773


Velocity m s 1Ecdyonurus sppVelocity ms

10j/>CO

■O>T3C

H—ocooz

1.8- 0.2 0 2 0.6 10 14

22

£OJcoXIo

10

oCOO

Z

14 1.8- 0.2 0.2 0 6 10Velocity m s "1 Velocity m s ~1

Ancylus fluviatilis Elmis aenea

oz

18

14

10

6

2

2 L_-0.2 1.00.2 0.6 1.4 1.8Velocity m s -1 Baetis rhodani

90V)(0=3 70"O>T3 50 c>*—o 30COoZ

-10-0.2 02 0.6 1.0 1.81.4

Velocity m s 1Ephemerella ignita

F ig u re 4 -6a , B lane W a t e r ve loci ty s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n of velocity a t s a m p l in g points . A b u n d a n c e values a r e n u m b e r o f in d iv id ua ls p e rsam p le .

109


151413121110

9876543210

CDCM CD CO CN CD

Velocity msr 1

22

18

14

x i 10

6

2

2 —-0 2 0.2 0.6 1.0 14 1.8Velcoity m s ~1

Limnius volckmari

4.5

3.5JLOCO13^ 2.5>X)c

o</)J? 0.5

-0.5-0 2 0.6 1.0 14 180.2Velocity m s 1

Agapetus fuscipes

30

26

w 22 (0I 18I 14 c

10H—ooz:

-0.2 0.2 0.6 10 14 18-1Velocity ms

Leuctra spp

16

14

1210864

202 —-0.2 0.2 1.00.6 1.4 1.8

Velocity m s 1Glossoma boltom

11

9

7

5

3

1

1 —-0.2 0.2 1.00.6 1.4 1.8

Velocity m s 1Oulimnius spp

F ig u re 4-6b , B lan e W a t e r veloci ty s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n o f velocity a t s a m p l in g po in ts . A b u n d a n c e va lues a r e n u m b e r o f in d iv id u a ls p e rsam p le .

110


4.3.4 Velocity

River Etive

The majority of the taxa (6 out o f 8) assessed were most abundant in samples where

velocity was 0 - 0.2ms'1, the most frequent velocity class, Fig 4-4. Of these

Oligochaeta, Tanypodinae and Tipulidae showed a decreasing maximum sample

abundance which is coincident with a decrease in occurrence o f velocity, Fig 4-1. The

totalled abundance of Hydroptila spp., Tanypodinae, Tipulidae and E. ignita were also

significantly correlated with the frequency intervals, Table 4-2. Baetis rhodani,

Hydroptila spp. and Ephemerella ignita appear to have a broader tolerance than the

other three taxa, occurring in relatively high numbers up to 0.5ms'1 or 0.7ms'1 in the

case of B. rhodani, Fig 4-4. The Orthocladiinae show maximum abundance in samples

of higher velocity 0.3-0.4 ms'1. Of all the taxa, Ephemerella ignita and the Hydroptila

spp. were the only taxa which have their maximal abundance correlated to frequency

of velocity intervals.

Duneaton Water

The frequency distribution of velocity at the Duneaton Water was log-normal and

similar to that at the R. Etive, Fig 4-5a, b. Few taxa exhibit the same pattern of

maximum sample abundance coincident with the commonest velocity observed at the

R. Etive. Only the maximum sample abundance of the Oligochaeta and Chironomidae

showed a significant correlation with the frequency of velocity intervals, Fig 4-1. The

totalled abundance of both these taxa were also significantly correlated with velocity

intervals, as are those of the Tipulidae and Oulimnius troglodytes, Table 4-4.

The maximum for E. ignita was near 0.25 ms'1 (excluding the outlier at 0.7 ms'1)

somewhat higher than at the R. Etive although it again exhibited a wide tolerance.

Baetis sp. Ancylus fluviatilis, Hydropsyche sp. and Limnius volckmari all have a

111


maximum abundance between 0.3 - 0.4 ms’1 which was higher than that at the most

abundant velocities. L. volckmari also appeared to have a wide tolerance. The

remaining taxon, Ecdyonurus spp. had its highest abundance at 0.2 ms’1.

Blane Water

The distribution of velocity is unimodal, skewed to the right and different from the

other two rivers, Fig 4-6 a,b. The most abundant velocity here was higher than at the

other two sites being between 0.2 and 0.4 ms'1. Here only Ecdyonurus spp. appeared

to follow the distribution o f velocities but, like all other taxa at the site, its maximum

sample abundance did not show a statistically significant correlation with frequency of

velocity. Its totalled abundance was significantly correlated with depth intervals as was

that of E. ignita, Oulimnius and Leuctra spp, Table 4-4.

The most frequent velocity interval was around 0.6 - 0.7 ms’1. The majority of the taxa

appeared to have their maximum sample abundance at slightly higher velocities than

this interval; included in this list are Elmis aenea, Glossoma boltoni, Leuctra spp.,

Oulimnius spp. Agapetus fuscipes, Ancylus fluviatilis and Baetis rhodani, Fig 4-6 a,b.

L. volckmari had an even higher maximum sample abundance at 1.0 ms’1.

Cross site comparisons

Taxa occurring at all sites

E. ignita

At the Blane Water and the Duneaton Water, E. ignita occurred in high numbers in

samples at all available velocities. At the R. Etive there was a sharp decrease in

occurrence around 0.5 ms’1. . At the other two sites where it was more abundant, it

occurs at higher velocities, 0.7 ms'1 in the Duneaton Water and 1 ms’1 in the Blane

Water. The low abundance at these velocities in the R. Etive would appear to be a

112


function of the very low abundance of E. igttita in the upland river as opposed to

reflecting any particular velocity preference.

So although significant correlations between maximum sample abundance of E. ignita

and velocity were recorded for the R. Etive data, the response to velocity should be

viewed cautiously. The Blane Water had the widest range of velocities and high

abundances of the animal and should give a truer reflection of the animals preferences.

Here the totalled abundance of animals was only correlated with velocity intervals.

Baetis spp.

At both the Duneaton and Blane the taxon showed maximum abundance in samples

with velocities just higher (0.4 and 0.5 ms'1 respectively) than the most available

velocities; 0.1-0.2 ms'1 and 0.2-0.4 ms'1 respectively. It was not correlated with

velocity in any of the rivers.

Taxa occurring in the R. Etive and Duneaton Water

At both sites the Chironomidae showed a preference for low velocities although there

is one outlier at 0.7 ms'1 at the Duneaton Water. At both sites, totalled abundance was

significantly correlated with depth intervals as was sample maximum at the Duneaton

Water. Abundance of Oligochaeta at both sites was low and decreased with increasing

velocity in a linear manner. At the Duneaton Water it showed significant relationships

with available velocity intervals. The Tipulidae exhibited a wide tolerance for all

velocities encountered.

Taxa occurring in the Duneaton Water and Blane Water

Leuctra spp. occurred at high abundances across all velocities with the animal

exhibiting a wider range in the Blane.

113


18

16

(/)coTO £ d) .9(0 ,1 ] (.2,.3] ( 4. 5] (.6,.7] (.8, 9]

Depth m

5.5

4.5J/)? 3.5"D>xi 2 5 c

o(/)oZ

-0.5-0 1 0.1 0.3 0 .5 0.7 0 9 1.1

Depth mEphemerella ignita

7

65

4

xi 3

2101 ■ ■ ■ ------0.1 0.1 0.3 0.5 0.7 0 .9 1.1

Depth mBaetis rhodam

1210864

202 . ------0.1 0.1 0.3 0.5 0.7 0 .9 1.1

Depth mHydroptila

7

65

4

3

2101 -------------------------0.1 0.1 0 3 0 .5 0.7 0 .9 1.1

22

18

14

10

6

2

2 -------------------------------------0.1 0.1 0.3 0 .5 0.7 0.9 1.1

Depth m Depth mChironominae Limnlus volckmari

F ig u re 4-7a , R iv e r E tivc d e p th s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n of d e p th at s a m p l in g po in ts . A b u n d a n c e va lues a r e n u m b e r o f in d iv id u a ls p e rsam ple .

114


X)o

18

16

14

1210864

20

< = 0 (.1 ,.2] (.3, 4] (.5, 6] (.7, 8] > .9(0 ,1 ] (.2,.3] (.4,.5] (.6, 7] (.8,.9]

Depth m

TO=3T?

0.1 0.3 0.5 0.7 0.9 1.1

Depth m Tanypodinae

3 5

3 0

w 2.5TO2.0

>T3 1 5 cOg 0.5 Z 0 0

-0.5-0.1 0.1 0.3 0.5 0.7 0.9 1.1

22

w3 14 ~o-o 10c:H—o

-0 1 0.1 0.3 0 5 0.7 0 .9 1.1

Depth m Depth mOligochaeta Tipulidae

1210864

202 . ------0.1 0.1 0.3 0.5 0.7 0 .9 1.1

Depth m Orthocladiinae

F ig u re 4-7b , R iv e r E tivc d e p th s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n of d ep th a t s a m p l in g po in ts . A b u n d a n c e va lues a r c n u m b e r o f in d iv idu a ls p e rsam ple .


(/>.2 10 (0« 8 inja° R'omo2

<=0 (1 3, > 5

7

65

4

3

210

_1 . . .-0.05 0 15 0 35 0.55

(0, 1] (.2, 3] (4 . 5]

Depth m

0 05 0 25 0 45

Depth m Ancylus fluviatilis

3 5

3 0

2.5

o

00-0.5

-005 0 15 0 3 5 C0 05 0 25 0 45

55

Depth m Anabolia nervosa

16

14

v> 10 (0-D 8>-O 6oino2

-0 05 0 15 0 35 0 550 05 0 25

Depth m Baetis sp.

65

55

45inco<0£o$ 25

X )oomo2

V-0 05 0 15 0 35 0 55

0 05 0 25 0 45

Depth m Chironomidae spp

70

60

50

ro 40

20

-101 ................-0 05 0 15 0 35 0 55

0 05 0 25 0 45

Depth m Ecdyonums sp.

45

35

25

15

5

5 L , . .-0 05 0 15 0 35 0 55

0 05 0 25 0 45

Depth m Ephemerella ignita

F ig u re 4-Sa, D un ca to n W a t e r d e p th s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n o f d e p th a t s a m p l in g points. A b u n d a n c e values a r e n u m b e r o f in d iv idu a ls p e rsam ple .

I 16


0 (0, 1] (1 . 2] ( 2, 3] ( 3, 4] ( 4, 5] > 5

D epth m

4 5

3 5

2 5

1.5

0 5

-0 5-0 05 0 05 0 15 0 25 0 35 0 45 0 55

Depth m Otigochaeta sp

05 0 05 0 15 0 25 0 35 0 45 0 55

Depth mHydropsyche sp.

0 05 0 15 0 25 0 35 0 45 0 55

Depth m Ouhmnius tuberculatus

28

'oino

-0 05 0 05 0 15 0 25 0 35 0 45 0 55

Depth m Leuctra spp

18

14

1062-2-0 05 0 05 0 15 0 25 0 35 0 45 0 55

Depth m Tipulidae spp

876543210-1-0 05 0 05 0 15 0 25 0 35 0 45 0 55

Depth m Limnius volckmah

Figure 4-8b, Duneaton W ater depth scatter plots with a histogram showing the frequencydistribution of depth at sam pling points. Abundance values arc number of individuals persample.

117

999999999999999458


121110

9876543210

< = 0 (.05,.1] (.15,.2] (.25,.3] > 35(0, 05] ( 1 .15 ] (.2,.25] (.3,.35]

Depth m

ro3"O>

18

14

10

6

2

-2 ■ . —-0 .05 0 05 0 .15 0.25 0.35

0 00 0 .10 0 20 0 .30 0 40

Depth m Baetis rhodani

T3>

-0 05 0.05 0 15 0 25 0.350 .00 0.10 0.20 0 .30 0 4 0

Depth mAgaptus fuscipes

26

22inro3~o>-OcoCDo

0 .15 0.25-0.05 0.05 0.350.00 0.10 0.20 0.30

Depth m Ecdyonurus spp

0.40

22

H—o(/)oz

-0.05 0.05 0.15 0.25 0.350 .00 0.10 0 .20 0 3 0

Depth m Ancylus fluviatilis

0.40

>T3

864

20-2 —-0.05 0.05 0.25 0.350.15

0 .00 0.10 0.20 0 .30 0.40

Depth m Elmis aenea

F ig u re 4 -9a , B lan e W a t e r d e p th s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n of d e p th a t s a m p l in g points . A b u n d a n c e v a lu es a r e n u m b e r o f in d iv id ua ls p e rsam ple .

1 18


10C/>aoIS£ <1) CO X2o'oo

> .35.05, .1 15, .25,<= 0(0,.05) ( 1 ,15] (.2,.25] (.3,.35]

Depth m

22

3~o>T3C

oCOo

-0.05 0 05 0 .15 0 .25 0 350.00 0.10 0 .20 0.30

Depth mLimnius volckmari

0.40

110

•90 •

•CO • •ro3X>

70 •

> 50 •T5C .* * • •*—o 30 • ; • • •(O 10

• • • so - - • •

-10; • • •• • •

-0.05 0.05 0 .15 0 .25 0 350 00 0 10 0 .20 0 30 0.40

Depth mEphemerella ignita

30

26

in 22 ro=> 18 T3> 14 ~o- 10

0 35-0.05 0.05 0.15 0.250 .00 0 10 0.20 0 .30 0.40

Depth mLeuctra spp.

864

2o

-2 ■—0 05 0 .05 0.15 0 25 0.35

0 0 0 0.10 0 .20 0 .30 0.40

Depth m Glossoma boltoni

~o>x>

119

7

5

3

1-1 ■-0.05 0.05 0.350.15 0.25

0 .00 0 .10 0 .20 0.30

Depth m Oulimnius spp

0.40

Figure 4-9b, Blane W ater depth scatter plots. Abundance values are number of individuals persample.

4.3.5 Depth

1 19


River Etive

Of the three sites the R. Etive had the deepest sampling points extending down to 0.9

m, Fig 4-7a,b. Unfortunately because species occurred at very low abundances at the

site, with some exceptions (Hydroptila spp. and Chironomidae), caution must be

exercised in interpreting taxon responses. The majority of taxa (7 o f 9) occurred in the

highest numbers in samples which had the most frequently occurring depths, Fig 4-7

a,b. These taxa also tend to decrease in number mirroring the decrease in the

availability of depths. Taxa exhibiting this type of response include, Hydroptila, Baetis

rhodani, the Oligochaeta, Orthocladiinae and the Tipulidae. The maximum sample

abundance of Hydroptila and Orthocladiinae were significantly correlated with

available depths (Fig 4-2) while the totalled abundance of B. rhodani, Orthocladiinae

and Tipulidae were also significantly correlated, Table 4-4. Both the Tanypodinae and

the Chironomidae appeared to exhibit a preference for depths (0.4 m) just slightly

greater than the most abundant ones. Both taxa were present in low numbers and

caution should be exercised in interpreting these results.

Duneaton Water

The range of depths in the Duneaton Water was more limited than in the R. Etive

extending to 0.5 m, only Fig 4-8 a,b. All taxa followed the distribution of available

depths occurring at maximum densities at samples which had the most frequent depths,

of these Baetis spp., Limnius volckmari, Oulimnius troglodytes and the Tipulidae was

the only taxon whose maximum sample abundance was significantly correlated with

the frequency of depth classes Fig 4-2. However the totalled abundance of

Chironomidae, E.ignita, the Oligochaeta, Oulimnius tuberculatus and the Tipulidae

were correlated with available depths Table 4-4. Ancylus fluviatilis, Anabolia nervosa

and Hydropsyche sp. were not statistically assessed as they were deemed to have too

120


few occurrences. A. nervosa was the only taxon not to show a preference for deeper

water, but, again this species occurred in low numbers so the result should be

interpreted with caution.

Blane Water

Depth range at this site was even more limited than at the other two sites getting no

deeper than 0.35 m, Fig 4-9a,b. The totalled abundance o f Ecdyonurus, E. Ignita, L.

volckmari, Leuctra and Oulimnius were all significantly correlated with available

depths, Table 4-4. The maximum abundance of few taxa follow the available depths

directly in the manner of those at the Duneaton Water. The only taxa exhibiting this

kind of response are Ecdyonurus spp., Elmis aenea and Oulimnius spp. Of these the

maximum sample abundance of Oidimnius was the only one significantly correlated

with depths, Fig 4-2.

The majority (7 of 10) of the other taxa did exhibit a similar response occurring in the

greatest numbers in samples of depths between 0.15 - 0.20 m which are very common

depths in this data set, but just higher than the most abundant depths of 0.05-0.1 m.

Taxa exhibiting this kind of response include E. ignita, G. boltoni, L. volckmari, A.

fuscipes, A. fluviatilis and B. rhodani. Of these, E. ignita showed a preference for the

deeper sections. Leuctra is the only taxon to show a definite preference for the deeper

sections reaching maximum numbers at 0.35 m.

Cross-site comparisons

Taxa occurring at all sites

E. ignita sample abundance showed no significant relation to the frequency of depth

intervals at any site, but its totalled abundance did at the Duneaton Water. In both the

Duneaton Water and the R. Blane the animal appeared to follow the available depths

121


although at the R. Etive it showed a preference for deeper water. Once again numbers

in the R. Etive are low and were therefore not subject to statistical analysis.

Limnius spp. at the Duneaton Water its maximum sample abundance was correlated

with depth intervals, but its totalled abundance was not. This situation was reversed at

the Blane Water.

Baetis spp. sample maximum was correlated with depth intervals at the R. Etive and

the Blane Water and totalled abundance was correlated with depth intervals at the R.

Etive too.

Duneaton Water and the R. Etive

Only at the Duneaton Water did Oligochaeta totalled abundance correlate with depth.

A similar response pattern occurs at the R.Etive but was not significant.

At both the R. Etive and the Duneaton Water the totalled abundance of Tipulidae was

significantly correlated with depth. Maximum sample abundance was also correlated

with depth at the Duneaton Water.

122


-20 40 100 160 220 280

S ubstra te A verage length mmChironominae

< = 0 (50,100] (150,200] > 2 5 0(0,50] (100,150] (200,250]

Substrate Average length mm

TD>

-20 40 100 160 220 280

S ubstrate A verage length mmBaetis rhodani

5.5

4.5

"O>~o 2.5

1.5

0.5

-0.5• • •• •

-20 40 100 160 220 280

S ubstra te A verage length mmEphemerella ignita

-20 40 100 160 220 280

S ubstrate A verage length mmCaenis rivulorum

10inCO

oino

-20 40 160 220 280100S ubstrate A verage length mm

Hydroptila

F ig u re 4 -10a , R iv e r E tive s u b s t r a t e s c a t t e r plots w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n of s u b s t r a t e a v e ra g e leng th a t s a m p l in g points. A b u n d a n c e va lues a r e n u m b e r o fin d iv id u a ls p e r sam ple .

123


131211109876543210

< = 0 (50,100] (150,200] > 2 5 0(0,50] (100,150] (200,250]


-20 40 100 160 220 280

S u b stra te A vera g e length mm

Orthocladiinae

22

18

ro 14

T3 1 0

3.5

3.0 ••

</) 2.5ro-§ 2 .0 • • t o>TJ 1.5c- 1.0 • • • • • • •

g 0 .5•• • • • • 2

• • • • • • 0.0 • • • • • • • • mm m • •

• • • 1 • t • t • • • •

-0.520 40 100 160 220 280

S u b strate A verage length mm

Limnius volckmah

-20 40 100 160 220 280


Tanypodinae

•

3.5

3.0 •

to 2.505-g 2.0 •>

^ 1-5c

• •- 1.0 • • • • • • • m • • •

8 0.5• •• • z :• • • • • 0.0 • • • • • • • • • • m • •

• • • • • • • • «fc» • m m •-0.5

-20 40 100 160 220 280


Oligochaeta

-20 40 100 160 220 280


Tipulidae

Figure 4-10b, River Etive substrate scatter plots with a histogram showing the frequency distribution of substrate average length at sampling points. Abundance values are number of individuals per sample.

124


14

oON


70 60

§ 50 ^ 40 ^ 30w 20° 10 <no o

-10-20 20 60 100 140 180 220 260

Substrate Average length mm Ecdyonurus spp.

7654321

0-1

-20 20 60 100 140 180 220 260

Substrate Average length mm Ancylus fluviatilis

:> 25

-20 20 60 100 140 180 220 260

Susbtrate Average length mm Ephemerella ignita

1614

121086420

-2-20 20 60 100 140 180 220 260

Substrate Average length mm Baetis sp.

-20 20 60 100 140 180 220 260

Substrate Average depth mm Hydropsyche sp

655545

352515

5

-5-20 20 60 100 140 180 220 260

Substrate Average length mm Chironomidae spp

F ig u re 4-1 l a ,D u n e a to n W a t e r s u b s t r a t e s c a t t e r p lo ts w ith a h is to g ra m sh ow in g th e f r e q u e n c yd is t r ib u t io n of s u b s t r a t e a v e ra g e len g th a t s a m p l in g poin ts . A b u n d a n c e v a lues a r e n u m b e r o fin d iv id u a ls p e r sam ple .


1413121110

9876543210


4.5

<5 35roa^ 2.5-5c>♦—oto2 0.5

-0.520 20 60 100 140 180 220 260

S ubstrate A verage length mmOligochaeta spp.

>X >

28

22

16

10

4

220 20 60 100 140 180 220 260

S ubstrate A verage length mmLeuctra sp

>Xj

O

8

7

6

5

4

3

2

1

01•20 20 60 100 140 180 220 260

S ubstra te A verage length mmOulimnius tuberculatus

8

7

6

5

4

3

2

1

01

20 20 60 100 140 180 220 260

S ubstrate A verage length mmLimnius volckmari

18

14

10

6

2

•2-20 20 60 100 140 180 220 260

S ubstra te A verage length mmTipulidae spp

F ig u re 4-1 lb ,D u n e a to n W a t e r s u b s t r a t e s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n o f s u b s t r a t e a v e ra g e len g th at s a m p l in g poin ts . A b u n d a n c e va lu es a r e n u m b e r ofin d iv id u a ls p e r sam ple .

o o o o o o o o o o o o o S- r ( M ' t ( D ( O O M ^ ( O O O O N ^ r j° o o o o As ^ ^ f N - ^ r t D o o o o o o o o

O N ' f IO ffl O NW ^ - T - T - T - T - t N ( N

126


<=0 (10,20] (30,40] (50,60] (70,80] >90(0,10] (20,30] (40,50] (60,70] (80,90]

Substrate Average length m

m 22

> 14

70 11030 50 90-10 10

Substrate Average length mm Ecdyonurus spp

— 35<U 031 , 5 T3C , ,

oin 0 5oZ!

-0 530 50 70 90 110-10 10

Substrate Average length mm Agaptus fuscipes

• •• • • •

• • m m •

-10 10 30 50 70 90

Substrate Average length mm Elmis aenea

110

22

18

14

10

6

2

-25010 10 30 70 90 110

Substrate Average length mmAncylus fluviatilis

110joTO3■g>*DcoinoZ

-1010 30 50 70 90 110-10

Substrate Average length mm Ephemerella ignita

18

14

10

6

2

230 50 70 90 110-10 10

Substrate Average length mm Baetis rhodani

Figure 4-12a,B lane W ater substrate scatter plots with a histogram showing the frequencydistribution of substrate average length at sam pling points. Abundance values are number ofindividuals per sample.

127


131211109876543210

oii

Substrate Average length m

30

26

w 22 ro~o 18>o 14c

i/>oZ

-10 10 30 50 70 90 110

S ubstra te A verage length mmLeuctra spp

-o>

16

14

12

10

8

6

4

2

0■2

-10 10 30 50 70 90 110

S ubstrate A verage length mmGlossoma boltonl

22

U)5 14T3>■D 10cotf)oz:

-10 10 30 50 70 90 110

S ubstra te A verage length mmLimnius volckmari

2.2

° 0.6</)O2 0.2

-0 2-10 10 30 50 70 90 110

S ubstrate A verage length mmHydropsyche si It a! at

-O>X)

11

9

7

5

3

1

1-10 10 30 50 70 90 110

S ubstra te A verage length mmOulimnius spp

F ig u re 4 -1 2 b ,B la n e W a t e r s u b s t r a t e s c a t t e r p lo ts w ith a h is to g ra m sh o w in g th e f r e q u e n c yd is t r ib u t io n o f s u b s t r a t e a v e ra g e leng th a t sam p l in g poin ts . A b u n d a n c e v a lu es a r e n u m b e r ofin d iv id u a ls p e r sam ple .

128


4.3.6 Substrate

River Etive

The distribution of substrate in the R. Etive was bimodal reflecting the high proportion

of both fine substrates and, at the other end of the scale, boulders and bed rock, Fig 4-

10a,b. Only the maximum sample abundance of Hydroptila spp. was significantly

correlated with substrate intervals at this site, Fig 4-3. The highest abundances of other

taxa are predominantly at the lower end of the substrate size scale. The Orthocladiinae

were the only exception having two samples with relatively high abundances in

samples dominated by larger substrate elements. The totalled abundance of none of the

taxa present showed a significant correlation to substrate intervals.

Duneaton Water

Availability of substrate lengths decreased with increasing length of substrate elements,

Fig 4-11 a,b. The maximum sample abundance of three taxa was correlated with the

distribution of substrate intervals; Chironomidae, Oligochaeta and Tipulidae. For the

last taxon the relationship was very strong, Fig 4-3. The other taxa show a range of

responses; the maximum abundance of Leuctra spp. for example appeared to be

inversely related to average substrate length, but this relationship was not significant.

Oulimnius tuberculatus occurred predominantly in finer substrates while Baetis spp.,

Ancylus fluviatilis, Limnius volckmari, Ephemerella ignita and Hydropsyche larvae all

appeared most abundant at around an average substrate length o f 100 to 120 mm. The

totalled abundance of none of the taxa present showed a significant correlation to

substrate intervals.

Blane Water

The distribution of substrate was unimodal and therefore different from that at the

other two sites, Fig 4-12 a,b. The range of average substrate lengths was much less

129


here reaching only 90 mm. A high proportion of the taxa exhibited unimodal type

responses but only the maximum sample abundance of Leuctra spp., Limnius

volckmari and Oulimnius spp. were significantly correlated with the availability of

substrate intervals, Fig 4-3. These 3 taxa plus B. rhodani and E. ignita had their

totalled abundance related to the distribution of substrate intervals, Fig 4-4.

4.4 Discussion

Despite the limited success of fitting Gaussian response curves to the data it is clear

that many of the taxa exhibit unimodal or linear responses to the data. Where curve

fitting was possible some of the taxa exhibited what appears to be bimodal

distributions. This is likely to be a sampling artifact as the lower velocities and depths

were sampled more frequently than the higher velocities / deeper depths, increasing the

chances of animals being sampled there. Given that the higher velocities and deeper

depths are less frequent and these taxa occur in relatively large numbers at these points

suggests that these areas are actively sought out, e.g. it is likely that Oulmnius

tuberculatus has a preference for faster waters.

When the taxa exhibit a maximum density in relation to a flow variable they frequently

occur at lower densities at the same measure of that variable, i.e. in the Blane Water

Limnius volckmari has a maximum density of 18 at a depth of 15 cm but it also occurs

at a density of 1 at that depth. This factor would have reduced the correlation

coefficients of the regressions.

As others have used alternative measures of abundance, including biomass (Makipaa

1999) or even using presence/absence in combination with logit regression (Saetersdal

& Birks 1997) for other groups suggest a certain flexibility in approach. I would

therefore advocate the use of maximal or summed abundance values, depending on the

question being asked, as possible alternative values.

130


The responses o f individual species indicate a close correlation with the distribution of

the environmental variables measured. What is clear from the results is that often when

an animal is dependent on the distribution of the environmental variable its maximum

sample abundance is too. This suggests that the taxon is either adapted to the ambient

conditions at the time of sampling or there is a statistical relationship between the

chance of getting a sample with a high number of individuals in it and the number of

samples at a particular point along the environmental gradient. Given the dynamic

nature of the variables measured, particularly velocity and depth it would seem

unlikely that the animals are that specifically adapted to the ambient conditions at the

time of sampling. That most of the taxa have a broad tolerance for depth and mean

water column velocity would seem most likely. Given that some of the same taxa

exhibit this kind of response at different sites to different variables over different

sections of the gradient supports the case for a statistical artefact. This is particularly

obvious at the R. Etive where abundance was low compared to the other sites. The

maximum sample abundance of some taxa showed strong correlations to the

distribution of velocity intervals. When compared to the other river where the species

occurred (the Blane) it was clear that the animals were within their natural tolerance

range and could reach much higher abundance.

It would seem advisable in future work to have a non-random sampling strategy and

take equal numbers of samples along the environmental gradient in question. Where a

stratified sampling regime has been used it was more successful at identifying

microdistributions of Hydrobiosidae larvae than the results shown here (Collier et al.

1995). The stratified sampling regime had been based on obtaining combinations o f

velocity and depths in the range of 0.1 - 1.5 ms'1 and 0.1 - 1.5 m respectively, using

five increments. Unfortunately some combinations o f velocity and depth were difficult


to obtain leading to ‘over sampling’ of some combinations. This is a situation likely to

occur in Scotland too. However, retaining the use of maximum sample values and

relating them to intervals of the environmental gradient would help to control for any

bias in the sampling.

This chapter is not meant to suggest that general survey work should not be carried or

that it cannot be used to identify species-habitat relationships, rather that care should

be taken in interpreting the results of such survey work in a quantitative manner, e.g.

CCA analysis has the capacity to identify relationships between organisms and

environmental variables even with this type of data. This is particularly the case for

PHABSIM studies were the results can be used to control discharge from reservoirs.

There is not only the chance that the incorrect ecological flows will be released but

that by releasing too much water the reservoir managers will lose money; water is

money.

One alternative option for survey work is to concentrate on a single or few species.

This type of study can be very successful at identifying the spatial distribution of

benthic invertebrates and related flow preferences (Cudney & Wallace 1980; Waringer

1987).

4.5 Conclusions

• Data collected using Instream Flow Incremental Methodology on benthic

invertebrates should not be used to model available hydraulic habitat.

• Structured sampling regimes need to be used if species flow preference curves are

to be recorded correctly.

132

Flume experiments

Chapter 5; Flume experiments; entrainment velocities of benthic

invertebrates

5.1 Introduction

The expression of an animal's tolerance of an environmental variable as being bounded

by a minimum, optimum and maximum is not only intrinsically satisfying but also

useful in the field o f water quality management and restoration projects (Peeters &

Gardeniers 1998). Water movement is a fundamental environmental variable in lotic

systems and can be viewed as the main engineer of the system habitats (Petts et al.

1995). Of the various aspects o f water movement velocity is the most temporally

variable. The aim of the work reported in this chapter was to accurately measure the

capacity of different invertebrates to resist entrainment. The point at which this occurs

can be viewed as the animal’s upper tolerance limit for velocity.

Of the three measures, the upper tolerance was chosen as the most useful. Managers

controlling regulated rivers frequently need to release water from dams for

maintenance and water level control behind the dam wall. These releases presumably

exceed the upper flow tolerances o f macrobenthos because they are known to cause

considerable disturbance of community structure in the receiving channel (Dejalon &

Sanchez 1994). The lower tolerances o f the animals are likely to be provided by the

release of Q90 or Q95 flows, a SEPA requirement requested during the planning

process for the discharge below dams in Scotland (Town and Countryside Act 1997).

Determining the optimum velocity for stream benthos, although attractive, is difficult

to measure in the field and artificial channels, particularly for the more mobile taxa.

The capacity of benthic invertebrates to withstand high velocities has repeatedly

intrigued researchers. Some very early work had excellent observations on the

importance of shape and behaviour of aquatic insects in withstanding entrainment

Flume experiments

(Dodds & Hisaw 1924; Dodds & Hisaw 1925). When the upper velocity tolerances of

individuals were actually measured, variations in results between workers were

considerable (Dittmar 1955; Dorier & Vaillant 1955) leading Hynes (1972) to observe

that laboratory experiments were not particularly informative, probably due to the

unnatural conditions to which the animals were exposed. There are probably many

reasons for the inconsistencies observed, the most obvious being that the velocities

measured were not recorded in front of the animals, but further up in the water column

(Dittmar 1955; Dorier & Vaillant 1955). Other complicating factors include the fact

that the animals were on different substrates. Neither of Dittmar’s studies reported the

depth at which velocity measurements were made so it was impossible to calculate

Reynolds number (which would have allowed more direct comparisons). Although

some studies did calculate the force of water incident to the animals, the measurements

were not taken from in front of the animals. To compound the difficulty with

interpreting these results, some o f the experiments were carried out in circular flumes

where vortex forces apply (Bournaud 1963). The accuracy of the results is therefore

questionable.

An increased awareness of invertebrate drift stimulated large numbers of flume and

field experiments which began to accumulate the lengthy list of factors which can

influence drifting. Velocity and substrate are included among these factors

(Ciborowski 1983; Holomuzki 1996). The majority of flume experiments avoided

studying the responses of individuals looking at groups on animals instead, but did

report on the importance o f density; drift does increase with density. (Borchardt &

Statzner 1990; Ciborowski 1983; Ciborowski 1987)

Some experiments looked at drift under extreme flow conditions in flumes, but again

the interest was not in individual capabilities (Lancaster 1992). Rather, the interest was

134

Flume experiments

in the importance of disturbance events. Whilst it was clear that both natural and

human simulated spates had the capacity to cause animals to drift (Matthaei et al.

1997; Matthaei et al. 1996), it was not until later that it could be shown that animals

could avoid being washed out of the system by accumulating in refugia (Lancaster &

Hildrew 1993 a). Both active movement and passive drifting were observed (Lancaster

1999) which contributed to no net loss in animals from spate channels and control

channels if refugia were present. This capacity helps to explain the exceptionally fast

rates o f recolonisation that are observed post spates (Death 1996). The circumstantial

evidence was mounting that animals were not only being entrained passively, but

would actively enter the drift. Supporting evidence already existed since it had been

shown that baetids drifted in response to the distribution of food resources within

experimental areas in artificial channels (Kohler 1985).

An animal can choose where it is going but, by drifting, the animal would appear to

forfeit control of its choice. This view is repeatedly seen in the work that does

concentrate on individual animal responses to extreme flow conditions.

Laser Doppler Anemometry (Statzner & Holm 1982) proved that a laminar sub-layer

would be too thin to afford protection to invertebrates from turbulent flow, which was

contrary to the long held understanding that they could (Ambuhl 1959). It was

necessary, in light of this discovery, to explain how animals resisted entrainment.

Reviews prior to Statzner’s discovery concentrated on the anchoring equipment and

streamlining of animals (Hynes 1972). Now the concept of drag was introduced

(Statzner 1988). As Vogel (1994) states 'it is easy to define drag as the removal o f

momentum from a moving body by an immersed body' but 'where in it does shape

enter?'. Some animals are streamlined, others flattened, in attempts to reduce drag but

it has been pointed out that such adaptations can have counter-intuitive implications

135

Flume experiments

i.e. flattening can cause positive rather than negative lift which has the potential to lift

the animal off the substrate (Vogel 1994). As drag and lift are not the only factors

affecting the animals it became clear that any animal shape is a compromise (Statzner

& Holm 1989) and the need to acquire food also can conflict with the demand to

reduce drag (Hart 1991). Even the different instars have different ideal drag-reducing

shapes because they live at different Reynolds numbers (Statzner 1988). The need to

compromise means that either structural anchors (claws or suckers) and muscle power

must be used to avoid entrainment. Ecdyonurus can achieve negative lift as its legs act

to produce a downward force, but it is likely that it must remain static to achieve this

because any movement would mean altering the angle o f the legs presented to the on

coming water (Weissenberger et al. 1991). Any movement from that location would

require energy to be expended, unless the animal was to actively drift out of this

position. The attachment mechanism of water pennies does allow movement across

rocks exposed to high velocities with potentially limited effort, but this is likely to be a

rather exceptional case as this taxon spends most o f its time on the underside of stones

(McShaffrey & McCafferty 1987, Smith & Dartnall 1980).

A simple measure of the energy required to resist entrainment was calculated by

subtracting the animals' passive resistance to drift (when dead) from its active (live)

resistance (Waringer 1989b; Waringer 1993). This approach was not new: Walton

(1978) had also compared dead and live animals. This was achieved by accurately

measuring the animals’ frontal area exposed to oncoming flow and calculating the

velocity to which it was exposed across this area at entrainment. The relative

importance of the animals’ shape and muscle power were not assessed. However the

results of the analysis did predict the distribution of some of the animals in the field

(Bacher & Waringer 1996; Waringer 1989a). Waringer’s field data suggested

136

Flume experiments

Allogamus auricollis Pictet should never occur outwith areas where it could resist

flow passively, although this was not the case for other Trichoptera. This suggested

that the situation was complex and that apparently similar species could behave rather

differently in real systems.

I decided to attempt this experimental approach. It not only gathers data on the upper

velocity tolerance of the individual animal, but gives some idea of the energy it must

expend to avoid entrainment. This is done by calculating the force exerted on the

frontal area of the animal by the moving water. I wished to extend the observations to

groups other than the Trichoptera. In reality, my success was limited for a number of

reasons. The principal difficulty was in obtaining accurate measures of velocity

incident to the frontal area of the experimental animals. However some potentially

useful data were obtained, and are presented here.

The experiments presented here tried to repeat Waringer’s work, but use

Ephemeroptera, Crustacea and Diptera, the aim being the same to determine the

amount of effort the animals put into maintaining position and to find the highest

velocity to which they are tolerant. The morphometric measurements were made on

the animals, but are reported here for only one of the animals; Ecdyonurus because

problems with the methodology prevented the study being extended to other species,

(see section 5.2.1).

Due to difficulties measuring accurately the upper velocity tolerance limit, and the

animals’ apparent ability to drift over a range of velocities, a different approach using

logistic regression, was adopted to show the range of drifting velocities for the

animals. This formed the second aim of the chapter. Logisitic regression shows, in

graphical form, the probability of the animals drifting over a range of velocities, and

Flume experiments

was the method adopted here. The hypothesis being that with increased water velocity

the chances of the animals becoming entrained increases.

So to conclude the real aim of this chapter is to present some results salvaged from

attempts to replicate Waringer’s experiments. Morphometric relationships between

head width and head frontal area are presented for Ecdyonurus. Observations on the

behaviour of benthic invertebrates exposed to high water velocity made during the

experiments are presented. Some data is presented on the phototrophic responses of

the study animals (light was to be used as a non-invasive stimulus during the

experiments). Finally logistic response curves of Gammarus pulex and Tipulidae to

water velocity and the influence o f body weight on this response, for the Tipulidae.

5.2 Materials

5.2.1 Flumes

52.1.1 Waringer's Flume

A bench top flume designed and described by Waringer (1989b) was built but a

number of difficulties were encountered in accurately calculating velocity. Firstly, it

proved impossible to use conventional flow measuring techniques because water

depth was limited to < 0.05 m in most areas. The flow measurement technique used by

Waringer corrected surface velocity, measured with a piece o f floating polystyrene, to

velocity deeper in the water column i.e. in front of the animals.

To use this method at higher velocities a video camera was mounted above the flume

which recorded the movement of the polystyrene ball over a fixed distance marked on

the flume floor. On play back, the number of frames could be counted and since the

time length of each frame was already known, velocity could be calculated.

138

Flume experiments

To correct surface velocity to that occurring directly in front o f experimental animals

Waringer used diagrams relating water depth to velocity in Dingman (1984), not the

equations quoted in the paper (Waringer pers comm). These equations can only be

applied if the Reynolds number (Re) for a point in the flume can be calculated and this

is not possible from the data. Waringer calculated Re using the diagrams given by

Dingman (1984). However one cannot apply these diagrams unless the Re is known:

the argument seems uncomfortably circular. Secondly the diagrams can only be applied

to areas of steady flow. Steady flow describes flow where velocity at a point is

constant or varies in a predictable manner, all fluctuation in flow is at a molecular level

(Walker 1995). Flow in the flume constructed at Glasgow proved to be unsteady, due

to the creation of a hydraulic jump near the start of the channel. This led me to

reappraise the flow-measuring techniques used above. Flow was also unsteady because

the flume widens toward its end, so depth and velocity vary constantly. Alternative

methods of measuring flow including the use of dyes did not prove successful. An

alternative flume was required which had the capacity to produce velocities o f >1.5

ms'1, and permit the use o f standard electromagnetic velocity meters Such a flume is

described below.

139

Figure 5-1, R

owardennan

flume,

side elevation. Units

are m

. A, experimental channel, width

0.1; B, M

ain reservoir

width 0.5; C,

Sm

all reservoir

containing flow

homogenising

marbles, width

0.1; D, Gate

valve; E, Adjustable

height support and

F, flexible tubing.

Flume

experiments

Flume experiments

5.2.1.2 Rowardennan Flume

Most of the experiments were carried out in a narrow flume 0.1 m wide, (Figure 5-1).

The water circulatory system is closed. Water was aerated by falling from the

experimental channel into the main reservoir. This process was sufficient to keep the

water saturated with oxygen during all operations; dissolved oxygen was measured

using a field probe (Hanna Dissolved Oxygen meter 1997 model). The flume was

housed in an unheated shed at the University field Station loch lomondside, under low

light intensities, and at ambient temperature. Water was supplied from Loch Lomond

(pumped from c. 10 m below the surface and 200 m from the shore). The water was

untreated, but was filtered through a phytoplankton net before use. The operation of

the pump could increase temperature by 1-2°C per hour and, when this occurred, the

water was replaced.

Discharge to the experimental channel was produced by a pump working at 12 Is'1 and

controlled by a gate valve. This allowed a very fast increase in discharge if required.

The slope of the channel could be altered by adjusting its height. Flow was

homogenised through a bed of marbles placed in the small reservoir. Depth in the

channel was controlled by means o f 0.02 m high perspex blocks which could be slotted

into the end. A zooplankton net was placed over the mouth of the main reservoir to

collect any animals swept out of the channel during experiments.

5.2.2 Flow Structure in the Flume

Flow in the channel varied gradually along its length, which was too short for steady

uniform flow to develop (i.e. flow was not full developed: the boundary layer did not

extend to the surface, but only partially through the water column, e.g. to a depth of 3

cm above the floor near the end of the flume). Estimation of velocity at different

heights above the bed using velocity profiles required the identification of the height of

141

Flume experiments

the boundary layer. In the boundary layer, the log-normal relationship between

velocity and depth applies and can be used to estimate velocity at different depths, i.e.

in front of an animal. Above the boundary layer velocity, does not vary with depth so

no estimation process is necessary.

To calculate the force o f water acting on the surface of the animals we can apply

equation 6.32 from (Dingman 1984).

Where u is the velocity at depth y (See Appendix I for definitions of terms)

This equation applies to hydraulically smooth flow, as is the situation when contact is

made with the substrate. It was developed for wide channels rather than the narrow

flow and channel used in the present experiments. As a result, their assumptions for

This replaces Sf for S0, where S0 = Channel slope tangent.

Manning’s n was estimated as 0.01 for flume with Perspex sides.

Equation 6-33 in Dingman (1984) can be applied for rough flow but ks would have to

be calculated.

v

calculating U* are not valid. Instead, U* in gradually varying flow is

n = Mannings n for the channel

Sf = Slope of the energy line

U' = -JgRSf

Flume experiments

Both equations 6-32, and 6-33 in Dingman (1984) are for fully-developed flow but can

be applied to the case in question here since we are only interested in the region of

developed flow at the bottom of the water column (D.A. Ervine, Glasgow University

pers comm).

5.2.3 Animals

All o f the species used are common in Scotland and are sufficiently abundant in rivers

to provide numbers for experimental analysis. The mayfly taxa Baetis rhodani and,

Ecdyonurus, the dipteran Tipulidae, and two crustacean species Gammarus pulex,

and Asellus aquaticus were all used. It was not possible to identify the Ecdyonurus

larvae to species but they were either E. torrentis or E. dispar, both of which were

present at the site. During some of the later work adults emerged: all were E.

torrentis. The Tipulidae were all of one species as yet not identified as no keys are

available to identify this larvae to species. Specimens have been retained and it is

hoped to rear some larvae of the same species to adulthood this year. Keys do exist for

identifying adults. All animals were collected from the Blane Water (site description,

Chapter 2 Section 2). Animals were collected by a number of means, usually stones

were turned over by hand and the animals allowed to float downstream into a pond

net. Where animals remained on rocks they were gently removed either with the finger

or an artist's paint brush to prevent damage to their cerci and other appendages.

Animals were transported to the field station in containers of river water which were

aerated using a portable air pump. On arrival, they were immediately transferred to

communal tanks (0.3 x 0.2 x 0.15 m), one tank per taxon in the shed housing the

experimental flume, which were also constantly aerated. Water was changed in the

tanks every 2 -3 days. Animals were not fed during their captivity but none were kept

longer than a week. Mortalities were infrequent.

143

Flume experiments

5.3 Methods

5.3.1 Weighing and Morphometric Measurements

Animals were removed from their holding tanks, carefully dried on absorbent paper

then weighed. Subsequently they were observed under a binocular dissecting

microscope and the head width measured using an eye piece micrometer. Animals

were then transferred to individually numbered plastic cups and remained in the flume

shed until they were used in the experiments. Tipulidae lasted in this type of

confinement without any reduction in observable health for up to a week, Gammarus

pulex and Baetis rhodani remained in good condition for approximately 4-5 days.

For Ecdyonurus, head width was related to the frontal area that would be exposed to

the current. After measurements of head width had been made, the animals were then

killed and mounted on to pieces o f card which were covered in nail varnish. Animals

were posed on the card in a position assumed by Ecdynonurus when resisting

entrainment, as observed by Weissenberger et al. (1991). Animals were placed facing a

binocular microscope lens to which a video camera was attached which was, in turn,

linked to a computer, Figure (5-2).

144

Flume experiments

Tocomputer

Figure 5-2, Measurement of Ecdyonurus frontal area. Binocular dissecting microscope. A, Cast iron right angle mount for specimens with built in spirit level; B Video camera linked to computer.

An image of the animal was recorded using a frame grabbing video card and then

analysed using Sigma Plot software. The image of the animal was outlined by hand and

then filled: the irregular filled area could then be measured using the computer

145

Flume experiments

program. Pixels. Counts were transformed into SI units using a calibration based on

measurements taken with a slide micrometer.

5.3.2 Experimental design and Statistical Analysis

In all the experiments carried out, randomisation of treatment order and allocation of

individuals to treatment was used to avoid any experimental bias. Hurlbert (1984) has

suggested that randomisation procedures can inadvertently create patterns in the

designation of experimental subjects to particular exposures thereby creating a bias

that they were trying to avoid. After each randomisation procedure I checked to make

sure no aggregations occurred. The main factor of concern was particular exposures

aggregating at particular times of the day. All experiments were carried out in daylight

hours and stopped before the onset o f dusk.

Phototaxis:

All experimental taxa were first screened to determine their response to light. During

some of the experiments which follow, a quantifiable stimulus was given to the animals

to move upstream. Light was the only stimulus which would not interfere with water

velocity around the animal and could be applied to specific areas of the flume. To

determine the animals’ responses under non-flowing conditions, the experimental

channel was dammed at both ends and flooded to capacity. This exposure acted as a

control for later experiments. A board was placed over the top of the flume, 0.3 m

from its end and extended 0.5 m along the flume. Underneath this area the walls were

blocked with card. A bench lamp with a 60 W bulb was then placed on top of the

plank and shone toward the downstream end of the tank. This created a sharp border

between the light and dark zones of the flume. As mentioned earlier, the flume hut had

low background light levels. From groups of about twenty animals (depending on

availability) individuals were placed singularly in the flume, on the border line and

Flume experiments

exposed for 3 minutes to moving water, or, in the case of the control, still water.

Exposures were randomised as described above. Animals were placed parallel to the

border line facing neither into the light nor the shade. Head width and weight of all

animals was recorded. A non-parametric Mann Whitney U test was performed on the

data for the B. rhodani groups responding preferentially to either light or shade.

Statistical comparison between light and shade preferences under flowing water

conditions were not performed. When exposed to running water the vast majority of

the animals sought in the illuminated zone. Only a limited number of animals remained

in the dark zone. It was clear that light was of secondary importance to remaining in

the channel.

Rheotaxis and suimming

The initial aim of this work was not solely to observe the animals’ behavioural

responses to high water flow rates, rather it was to measure the upper velocity

tolerance of the various taxa. Animals were placed in the flume individually on the

border between the light and dark areas with the channel flooded and the water still.

They were allowed to settle for five minutes and then some of the barrier at the end of

the experimental channel was carefully removed in sections to prevent any surging

motion in the water upstream which would inadvertently entrain the animal. The

pump was then started and the gate valve, which had been kept completely shut, was

slowly opened which allowed water to begin flowing gently. Velocity was increased

slowly over 3 to 4 minutes by opening the gate valve further. When required, the

experimental channel was also angled upward to increase velocity. This was necessary

only in exceptional cases to achieve velocities over 1.2 ms'1 for ecdyonurids. When the

animal became entrained it was recovered from the end of the flume and, at the point

where entrainment occurred a velocity profile was measured at 10 positions through

147

Flume experiments

the water column, by taking 15s averages at each point. It soon became clear that the

animals had an impressive capacity to find areas o f the flume where velocity was low

(near the walls) as refuges and that these locations were also inaccessible to the

velocity meter. They also showed a willingness to drift over a range of velocities and

appeared unwilling to expend energy withstanding entrainment. For this reason it was

felt that measuring an upper velocity tolerance limit in this manner was not informative

and only the animals’ behavioural responses are reported.

Logistic response curves

The previous experiment determined the range of velocities at which the different taxa

normally drifted and this allowed a new experiment to be designed which

concentrated on determining the probability of animals drifting at different velocities.

The velocity range of animals had been determined in the previous experiment. From

the range a number of velocity ‘exposures’ were chosen (see Table 5-1). Animals were

collected, weighed and measured as described above. They were then assigned to each

exposure in a random manner, 5 animals to each exposure. Each animal was treated to

its velocity exposure individually for a period of 3 minutes and then its behaviour was

scored using the list of behavioural categories presented in Table 5-2. The order in

which each exposure was made was completely randomised.

Frontal area to head width

To predict the frontal area of Ecdyonurus exposed to flow from its head width a

simple linear model was used. Animals were chosen to cover the maximum range of

head widths available.

148

Flume experiments

Table 5-1 The velocity to which the three taxa were exposed.

Taxa Number of exposures Exposures ms'1

Tipulidae 7 0, 0.04, 0.06, 0.08, 0.1,0.12

Gammarus pulex 5 0, 0.10, 0.15, 0.20, 0.30

Baetis rhodani 4 0, 0.3, 0.5, 1.10

Table 5-2, Definitions of behaviour scored during exposures to velocity. Given in Table 5-1.

Behaviour Definition

Entrained Any movement downstreamResisting flow Struggling to remain in situ or moving to flume edgeActive Moving upstream or lateral movement without any downstream

movementInactive Remaining still

Effect o f weight on entrainment o f Tipulidae

Animals were collected in the field as previously described. Animals were ranked in

order of their weight. Of these, 30 animals were chosen covering the range of weights

available. This equated to 1 individual representing every 0.01 g increment in weight.

Each individual was exposed in a similar manner to a series of velocities: 0.04, 0.06,

0.07, 0.08, 0.10, and 0.12 ms'1. The order in which the animals were selected for the

experiments was randomised and the order an individual was exposed to the velocity

intervals was also randomised. Animals would frequently become entrained at

velocities below their upper velocity tolerance. The lowest and highest velocity at

which an animal was entrained were recorded to take account of this behaviour.

149

Flume experiments

5.4 Results

5.4.1 Head width to frontal area ratios

There was a positive significant relationship between head width squared and frontal

area of Ecdyonurus (p < 0.0001, F (1, 73) = 144.5, R2adj =0.65). Although the

relationship was significant, there was fairly substantial degree o f individual variation.

Frontal area = - 0.40 + 0.409(Head width)2.

CN

EoCO<D(0oj■*-*co

LL

0 2 4 6 8 10 12 14 16

Head width squared cn?

Figure 5-3, The relationship between head width and frontal area of Ecdyonurus.

5.4.2 Phototaxis

All the taxa tested except Baetis rhodani showed a strong preference for the shaded

area, Table 5-3. The almost 50-50 response to either light or shade would indicate that

Baetis rhodani has no preference. There was no significant difference in the weights of

B. rhodani individuals which preferred shade to those that preferred light.

When the animals were exposed to the same conditions with moving water they all

initially moved upstream into the shaded zone, including B. rhodani. However as

velocity increased, they moved into the lit zone where they would still resist

entrainment.

76543210

O QCP

150

Flume experiments

Table 5-3, Percentage of animals preferring the shaded area over the lit area.

TaxaEcdyonurus Baetis

rhodaniGammarus

pulexAsellus

aquaticusTipulidae

Mean Weight g

0.015 0.005 0.045 0.011 0.191

Standarddeviation

0.009 0.003 0.062 0.006 0.088

n 20 19 20 20 20Shade

% of animals90 52 95 85 80

5.4.3 Rheotaxis and swimming

Ecdyonurus

This species, like B. rhodani, swam in short bursts . The length o f the bursts decreased

with increasing velocity (personal observation only). It remained close to the substrate

for the majority of the time but on occasion would swim further up into the water

column. When velocity increased B. rhodani would either enter the drift or, if the

water velocity was not sufficient to wash the animal out of the experimental channel, it

would frequently alight on the floor and then enter the drift again. There was a marked

decrease in an individual’s upper velocity tolerance with repeated exposure making it

impossible to use the same animal for two consecutive runs. This was only attempted

if there had been a problem during the first run. On a number of occasions individuals

would resist entrainment until the velocity was very high >1.5 m s1. When this

occurred they assumed a flattened position with their body pressed closely to the floor.

This posture was taken up as velocity increased although only infrequently did the

animals persevere until velocities were very high (>1.5 ms'1). On other occasions the

animals would move to the flume edge as velocity increased, presumably gaining

advantage from the reduced water velocity at these locations.

151

Flume experiments

Baetis rhodani

When moving upstream or downstream, the animal swam in bursts coming back down

onto the floor periodically. Above a certain velocity the animal would not swim

downstream: instead it appeared to drift. After leaving the substrate it would move up

through the water column rather than drift near the floor. It was not clear whether this

action was voluntary or passive. Once in the water column the animal assumed a

posture where its abdomen was arched upward and its cerci aimed forward over its

head. After travelling in this manner the animal would float down through the water

column and grasp the floor with its claws. If the animal had not travelled far enough

(assuming the animal is making the decision about what distance to travel), while still

suspended it would flick its abdomen when descending which stopped the descent and

allowed further travel downstream. Some individuals would also move to the side of

the flume where they could resist entrainment more easily. It was clear that an

individual would choose whether or not to enter the drift, although there was always a

point at which velocity was sufficiently high that they could no longer remain in

contact with the substratum.

Gammarus pitlex

Like the previous two taxa, this species showed a range of responses to high

velocities. Unlike the other taxa, G. pulex would quickly move upstream once

introduced into the flume. It also differed in its use of the whole water column while

swimming, something only B. rhodani did as frequently. As velocity increased G.

pidex would either get washed out of the experimental channel or alternatively would

move down onto the flume floor. There it would hold on with its frontal appendages,

from either thoracic or head sections, while facing upstream. This animal would

frequently tilt its body upwards to the rear, from its point of anchorage, at an angle of

152

Flume experiments

as much as 45° or more. Depending on the individual and the exposure velocity, three

related responses were observed. Firstly individuals would release their hold and drift

downstream (after moving up into the water column), alternatively they would release

and move downstream pushed by the flow, but would remain close to the floor and

periodically re-attach and then get swept down again, sometimes simultaneously

moving to the flume wall. Their capacity to perform this action was extraordinary; they

could on occasion move at right angles to the direction o f flow, although more

normally there was some downstream drift involved also. This movement at right

angles could occur even at quite high velocities.

Asellus aquaticus

Animal availability in the wild decreased shortly after the initial phototaxtic test and no

behavioural observations could be made in running water.

Tipulidae

When moving upstream or against the current, the animals appeared to bite the

substrate with their mandibles and while so anchored would draw their abdomen

further forward. This activity was accompanied by moving the head from side to side.

If the current proved too strong the animal would release its grip and roll downstream

angling its body so it ended up at the flume wall. This process was completed in stages

with the animal holding on, reorienting, and then releasing until it reached its

destination. Once it reached the flume wall it expanded its body into the comer

between floor and wall. Here velocity was not measurable, but was noticeably less

than in the centre of the channel.

The increase in discharge necessary to dislodge a tipulid from this refuge was

substantial. If the water velocity was too fast and the animal could not reach the flume

edge it would release its grip and drift downstream, usually rolling over the flume

153

Flume experiments

bottom. Like the other taxa, Tipulidae that were repeatedly placed in the flume

showed an increased readiness to drift and the velocity at which they drifted decreased

over the duration of the experiment. One individual appeared to attach by means of

silk thread, to the flume floor during high flow. This left the animal floating above the

substrate (circa 1 cm) and gave it the capacity to remain in position at much higher

velocities than other Tipulidae. Some Tipulidae do spin silk from labial glands

although this capacity is limited to the aquatic and semi-aquatic forms, (G. Hancock,

Glasgow University pers comm.).

5.4.4 Entrainment logistic curves

Tipulidae

There was no significant difference in the mean weight of the groups assigned to each

velocity (ANOVA), see Table 5-4 and Figure 5-4. The change from active behaviour

to either resistance to entrainment or to be actually entrained, occurred over a short

velocity range (Figure 5-5). Few animals showed resistance behaviour and all were

entrained when water velocity had reached 0.12 ms'1.

Table 5-4 Tipulidae mean and standard deviation of weight for each experimental exposure tovelocity, n = 5.

Exposurems’1

0 0.04 0.06 0.08 0.1 0.12

Mean g 0.230 0.202 0.152 0.189 0.160 0.221Standarddeviation

0.089 0.080 0.030 0.092 0.051 0.070

154

Flume experiments

COc:o03£CDCO

_QO

«4—Oo

13121110

9876543210

<=.1 (.1,15] (.15,.2] (.2,.25] (.25,.3] (.3,.35] > .35

Weight g

Figure 5-4, Weight distribution of Tipulidae used in logistic regression experiments.

1.2

1.0

> , 0 8

0.6

0.4

0.2

0.0

_Q03

XIO

- 0.20.00 0.02 0.04 0.06

Velocity ms

0.08

- 10.10 0.12

Figure 5-5, Logistic regression analysis of Tipulidae behaviour showing the probability of the animal exhibiting either resistance to entrainment or the animal actually being entrained. Equation of the line is y = exp(-7.35+(89.69)*x)/(l+exp(-7.35+(89.69)*x)), Chi square = 22.7, pcO.0001.

89999999999

Flume experiments

(/)CoCD

£Q)(/)

_QO

M—Oo

109876543210

<=.01 (.01,.02] (.02,.03] (.03,.04] (.04,.05] > .0 5

Weight g

Figure 5-6, Weight distribution of Gammaruspulex used in logistic regression experiments.

Velocity mFigure 5-7, Logistic regression analysis of Gammarus pulex behaviour showing the probability of the animal exhibiting either resistance to entrainment or the animal actually being entrained. Equation of the line is y = exp(-5.12 +(45.6)*x)/(l+exp(-5.12 +(45.6 )*x)), Chi - square = 20.58,p<0.0001.

Gammants pulex

T here w as no significant difference in the w eights o f anim als assigned to each

exposu re (A N O V A ), see Table 5-5 and Figure 5-6. The range o f responses to velocity

w as g rea te r for G. pulex than for the T ipulidae, F igure 5-7

156

Flume experiments

Table 5-5, Gammarus pulex mean and standard deviation of weight for each experimental exposure to velocity.

Exposure ms'1 0 0.1 0.15. 0.2 0.3

n 5 4 5 5 4Mean g 0.027 0.029 0.023 0.027 0.027

Standard deviation 0.013 0.018 0.007 0.012 0.012

5.4.5 Effect o f weight on entrainment o f Tipulidae

All animals were entrained in a short band of velocities. There were significant positive

correlations between weight and the lowest and highest velocity at which animals

became entrained. One o f the individuals used was very large (0.763 g) and considered

an outlier. Removal of the outlier did not affect the significance of the relationship

between velocity and the lower velocity o f entrainment. The relationship between the

highest velocity of entrainment and velocity did become non-significant on its removal.

0.085

0.075

i^ 0.065

O 0.055 <D >

0.045

0.0350.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

W eight gFigure 5-8, The lowest velocity of entrainment exhibited by the Tipulidae against body weight. Spearman rank correlation, R = 0.56, p<0.005, n= 29. Equation of the line; y = 0.052+0.044*x.

O O

O O CO O O CD O

o oo cm o o o

o oo o

157

Flume experiments

0.11

0.10

0 .0 9

'(/) 0 .0 8 E>> 0 .07 Oo 0 .0 6d)

^ 0 .0 5

0 .0 4

0 .0 30 .0 0.1 0 .2 0 .3 0 .4 0 .5 0 .6 0 .7 0 .8 0 .9

Weight gFigure 5-9, The highest velocity of entrainment exhibited by the Tipulidae against body weight.

Spearman rank correlation, R = 0.40, p<0.05, n= 29. Equation of the line; y = 0.06+0.053*x.

5.5 Discussion

The observed responses of the benthic invertebrates are consistent with the hypothesis

that animals frequently choose to enter the drift. Similar behaviour to that observed

here has been interpreted as a reluctance to drift (Lancaster 1999) and this is not

contradicted, rather it is suggested that a choice is being made and that the animals can

become more willing to enter the drift if stressed. In none of the exposures made

during the present study were the animals given any refugia. Their capacity to find the

edge of the flume where velocity was low, highlights a highly developed rheosensory

capability, across all taxa examined. This capacity together with the ability to alter the

length of drift - by either rolling (tipulids) or flicking the body (Baetis rhodani) up

into the water column - may help explain the capacity of invertebrates to choose

substrate type directly from drifting (Walton Jr 1978).

The negative phototactic response of all subject animals, excluding B. rhodani

supports the use of substrate as shelter. Where the river is not shaded by vegetation

the substratum is the only place which is dark. By seeking such a location

158

Flume experiments

macroinvertebrates are choosing an area in which is presumably difficult for visual

predators to operate. Areas which are dark are likely to be deep in the substrate matrix

and should have lower shear stresses than substrate elements close to the bed surface,

however the hyporheic refuge theory has only been supported in part by field

experiments (Palmer et al. 1993b) but the question here is more concerned with

animals sheltering not from serious floods but from ‘normal’ flow conditions. Animals

in the substratum would require less energy to maintain position.

However both G. pulex and E. ignita have been shown to use woody debris as shelter

when velocity is increased, although G. pulex made more efficient use of the refuge

(Borchardt 1993). Although no light measurements were reported from amongst the

woody debris it is clear from their size that they would have had some shaded areas.

From the evidence presented here it is possible that the two stimuli, both light and

current could work in concert or independently. Further work which identifies which

stimulus is dominant would give a possible insight into the relative importance of

predator avoidance versus entrainment.

That no response to light was observed in B. rhodani also fits with our knowledge of

the animal as a good swimmer and suggests that it is more flexible in its response to

negative stimuli than the other taxa, e.g. predators. The situation is not simple though

as both B. rhodani and heptageniids nymphs are known to seek shelter among the

substrata; in this case it was suggested that they are avoiding predators (Huhta et al.

1995). Baetis species are known to show strong diel periodicity in drifting (Poff &

Ward 1991; Tikkanen et al. 1994), but it is not clear if this is regulated by light levels

or circadian rhythms. In this instance, if B. rhodani was shown to drift at night, the

lack of preference for light or dark conditions would be indirect evidence for an

internal clock. As drifting in Baetis can be very seasonal and local drift records are few

159

Flume experiments

so no direct conclusions on this subject can be made (Rincon & LobonCervia 1997),

but the possibility for future work exists. 1

The Tipulidae occur at higher velocities in the field than those at which they were

entrained in the flume (see Chapter 4 Section 4.3). This could be for two reasons.

Firstly the flume substrate afforded less grip than the substratum in the field, or,

second, the animal remains below the surface of the river bed for the majority o f the

time. As the tipulids could withstand only very low velocities, lower than those

expected near the river bed, they must avoid the substratum surface. This highlights

the fact that these organisms are exploiting the habitat in very different ways and

suggests that niche differentiation is occurring along these physical gradients.

The increased ability of larger Tipulidae to withstand entrainment is a situation

observed in other taxa. For some Trichoptera it has been suggested, that because of

this relationship the chances o f accidental dislodgement decrease with size (Otto

1976). That the surface area to volume ratio of a small object is much greater than

that of a similarly shaped large object is often quoted in texts on thermo-regulation. It

is well established that the surface area to volume ratio of a small object is much

greater than that o f a similar shaped large object (Schmidt-Nielson 1990). If we

assume that weight will increase in a linear manner with increasing volume, ignoring

for the moment the stepped size increases exhibited by insect instars, then the ratio of

frontal area to weight o f a small tipulid will be relatively greater than that of a larger

specimen. Therefore, there would be relatively less force exerted on a larger animal

per unit weight. In theory, therefore, a larger animal would need to expend less energy

to withstand entrainment. This should be applicable to other taxa also: the only limit

1 Some 24 hr drift samples were taken from the Blane Water but the sample size was small and the results inconclusive. Directly upstream of the drift nets was a very large population of Hydropsyche

160

Flume experiments

would be the individual's height increase. As the animal’s height increases it could

project too high into the boundary layer subjecting the animal to relatively higher

velocities. This would be dependent on the increase in velocity with depth which, in

shallow riffles, can be steep. This would provide one explanation for the ability of

large tipulids to withstand higher velocities since the relative increase in volume

equates to increased muscle. Whether resistance is a physical act or passive inertia

needs to be clarified. From the observations of the animals’ behaviour it seems likely

to be a combination of both.

The velocities at which the Tipulidae and G. pulex drifted will have been influenced by

the substrate in the flume bottom. It seems likely that this type of work will increase

where it is necessary to get exact measurements of the flow preferences of

invertebrates. It would appear likely, therefore, that further studies of this type are

required during which exact measurements of the flow preferences of different species

of invertebrates could be made. Standard substrates would be used to allow

comparisons with both field conditions and other experimental studies.

Giller & Malmquist (1998) have suggested that a pluralist approach to understanding

community structure is necessary; it is no longer sufficient to only argue about the

relative importance of biotic versus abiotic factors. Drift by benthic invertebrates

appears to be a general response to both types of factors (Brittain & Eikeland 1988;

Collier & Wakelin 1992; Kratz 1996). The data presented in this chapter add weight to

the argument that individuals actively choose to drift and that this response is plastic at

the level of the individual. Individual plasticity is therefore a useful measure of the

relative importance of the various factors affecting community structure.

---------------------larvae which could also have significantly altered the composition of the drift (Georgian & Thorp 1992).

161

Flume experiments

The process of drifting is likely to reduce the fitness of the animal. By this I mean the

actual drift itself not the process which is actually beneficial e.g. drifting from one food

patch to another is consuming time which could be spent feeding, but the process may

be necessary if food is to be found at all (Kohler 1985). So if the drifting behaviour of

an animal is altered its fitness, as measured by fecundity, is likely to be altered too.

5.6 Conclusions

• Negative phototactic responses o f benthic invertebrates agree with their known use

of substrate as a shelter.

• Benthic invertebrates have strong rheosensory responses.

• Entrance into the drift is a matter o f choice below certain upper velocity limits.

These limits are species-specific.

• There is a positive relationship between body size and the capacity of a tipulid

larvae to withstand entrainment.

162

Invertebrate community structure in Callitriche stagnalis patches

Chapter 6: Invertebrate hydraulic microhabitat and community structure in Callitriche stagnalis Scop, patches

6.1 Introduction

When macrophytes are included in rehabilitation schemes for channelled rivers they

can potentially provide a diverse habitat for invertebrates by interacting with local flow

conditions (Marshall & Westlake 1990; Newbury & Gaboury 1993). For practical

management purposes it is useful to know how much additional invertebrate diversity

and biomass may be supported by utilising macrophytes as part of an individual river

rehabilitation scheme, under specific flow conditions. I use the term diversity of

invertebrates to encompass taxon richness, abundance and evenness (the relative

number o f individuals o f each taxa). Previous work e.g. (Gregg & Rose 1982) has

demonstrated the extent to which macrophytes can alter the stream flow

microenvironment. Large-scale studies (e.g. Jenkins et al. 1984; Ormerod 1988) have

demonstrated how this increased habitat diversity in turn influences the occurrence of

invertebrate assemblages in vegetated and unvegetated parts of the river environment.

Sand-Jensen & Mebus (1996) showed that a range of flow conditions were present in

patches of the submerged macrophyte Callitriche cophocarpa. Other studies (Wright

1992); Jeppesen et al. (1984) found that invertebrate communities present in

macrophyte patches (including Callitriche) could differ substantially from those

occurring in unvegetated patches of river bed, having higher taxon richness, numerical

abundance and specifically high numbers of Chironomidae and Simuliidae. Woody

debris has been show to have increased species richness if its complexity is increased in

streams, suggesting increased habitat complexity as the mechanism behind high species

richness and abundance in plant patches too (O'Connor 1991).

163


The Blane Water is a small unregulated stream which supports a Callitriche-

dominated macrophyte flora typical of many small British rivers in which rehabilitation

schemes are being considered or implemented. The relationship between flow

conditions and invertebrate community structure within different parts of submerged

beds of Callitriche stagnalis Scop., and in adjacent unvegetated patches of river

substrate in the Blane Water was investigated. The results are considered in relation to

the value added to river rehabilitation schemes by the inclusion of macrophytes, in

terms o f increased invertebrate diversity support capacity.

The aims of the chapter are to show that different sections of Callitriche instagnalis

support different assemblages o f benthic invertebrates and that this may in part be

related to the flow conditions in and around the plants. Specifically that the outside of

the plants is an extreme environment for benthic invertebrate suitable for a limited

number of specialist taxa only.

6.2 Methods

6.2.1 Source o f Plants

Plants and substrate were collected from riffle sections o f the Blane Water, a gravel

bed river, at Blane Bridge (UK National Grid reference NS507852) in Scotland during

June 1998. The site is down stream from a trout fishery and is classified as organically

enriched by the Scottish Environmental Protection Agency, although water quality is

classified as high (Doughty & Maitland 1994). The only noticeable effects of

enrichment are higher numbers o f chironomids and annelids than would otherwise be

expected. Channel width in the sampling area ranged from 8-13 m, depth from 7-22

cm and velocity from 0 to 0.35 ms'1.

164


6.2.2 Measurements

Ten Callitriche stagnalis stands were chosen at random within a c.100 m stretch of

river. Water velocity and depth were measured in front, above, in the middle and at the

bottom of each stand using an electromagnetic velocity meter (SENSA). At the mid

channel side of each plant, velocity and depth were measured and substrate type was

visually estimated (% scale) using the following particle size scale: sand (0.06 mm-2

mm nominal diameter), fine gravel (2-10 mm), gravel (10-64 mm) and cobble (64-256

mm). A reference scale was carried in the field to confirm particle size when necessary.

Mean water column velocity measurements taken outwith the stands were measured at

0.4 depth from the river bed (Smith 1975) as 50 second averages.

The total area of bed occupied by each stand was estimated using a quadrat with 25

cm2 squares. Invertebrate samples were taken separately from the outer, middle and

root sections of the plants by placing a net downstream and trimming the desired

section of the stand with scissors and letting it float into the net. Sections were defined

using stand length and depth as follows; outer = outer 20%, mid = all other material

above substrate, root = material within the bed. When removing the roots all the

substrate which had been underneath the stand was disturbed and the invertebrates

collected into the net placed downstream. A Surber sample was then taken from the

mid channel point where substrate type had been assessed, 10 surber samples in total. I

estimated the near bed velocities at the Surber samples by using values which were

10% of those at 0.4 depth from bed, as suggested by Carling (1992).These samples,

and the sections of stands, were immediately placed in 70% alcohol to preserve the

invertebrates. All samples were sorted on white trays and animals identified to the

following taxonomic levels: Annelida to subclass; Diptera to family or genus and all

165


other taxa to genus or species. Other studies show larvae and adults of Coleoptera

have differing hydraulic habitat preferences (Degani et al. 1993) and so are treated

separately here.

6.2.3 Data manipulation

To compare densities of invertebrates between stand sections, abundance values were

standardised to estimated wet weight of stands. Because of the necessity of

immediately preserving invertebrate samples in alcohol, in the field, the plant material

in the samples could only be weighed after a storage period in 70% alcohol. It was

necessary to see if samples o f different weights were affected equally. The relationship

between dry weight post storage with wet weight prior to storage was linear for log-

log data as determined by regressing the wet weight of 24 control stand samples

against their dry weight.

LnWet weight = 2.59+0.92(Ln Dry weight), (R2 = 0.978)

This equation was used to transform dry weight values to wet weight for the

experimental plants.

6.2.4 Statistical analysis

Where direct comparisons between samples were made the Wilcoxon Matched-Pairs

Signed-Ranks Test was used, which allows comparison between related samples

(Sidney 1956). As either different sections of stands are compared or stands and

adjacent sections of substrate (Surber samples) it was considered that they represented

related, non-independent samples and that this test was the most applicable. Benthic

invertebrates can have clumped distributions and it is possible that this could occur

around some of the Callitriche stands; see Elliot (1977) for a comprehensive

assessment of the possible spatial dispersions of benthic invertebrates. It was believed

166

___________________________________Invertebrate community structure in Callitriche stagnalis patches

this could further influence the relatedness of spatially close samples and enhanced the

applicability of the Wilcoxon Matched-Pairs Signed-Ranks Test

167


6.3 Results

6.3.1 Taxa abundance and richness compared between plant and surber samples

Although composed of many individual plant fronds, all stands had a similar growth

form, being wider at the upstream end and tapering downstream, similar to a classic

streamlined strut as described by Vogel (1994) but wider at the upstream end and

generally more ragged. The area of stream bed occupied by stands was on average

467cm2 (s.d.+/- 290cm2). Substrate in the Surber samples was dominated by gravel (%

composition of gravel always >90%). The substrate underneath the Callitriche

stagnalis stands was mainly sand (% composition of sand always > 80%). Callitriche

stands had higher, and significantly different total taxa abundance (no. o f individuals of

all taxa) and taxa richness than neighbouring substrate samples (Wilcoxon Matched-

Pairs test p<0.01, N=10. Values standardised for sample area: the area of bed

occupied by each stand. P value is for both cases, see Table 6-1 for abundance and

richness data per sample). A list of taxa occurring in the Surber samples and plant

stands is given in Table 6-2 with abundance (no. o f individuals per sample) and

frequency of occurrence (no. of samples a taxon occurs in).

6.3.2 Taxa abundance and richness compared between plant sections

Taxa richness was significantly different between outer and root sections of the stands.

From the box plot in Figure 6-1 it is clear that the outer section supports the lowest

taxa richness. Total taxon abundance was significantly different between outer and

mid sections and outer and root sections. The box plot shows that the outer section of

the stand supported the greatest abundance, see Figure 6-2, Table 6-1 for abundance

and richness data per sample.

168

1.29 1.81

1.99 546

90 0.14

0.24 0.49

ON

® * - »3 < a 35*S ' <W o- sr32 O c« "B•a *t32 S'

< $ P SOs 2

S» 1O I—b •

" E? £ ® a 2

NO 00 - J On On 4 u> t o C/3P3

* 2aT

ONtoC7NO3kl

MM-3p<-K2 #

ffc 3*a sS' 3 3 p<» 3

i—iNO

U) OnU)

OnO

t oNO O

U)On

l—ll—l OnOC

O>1

2Sm' 2 «> 3 «n » B CB 3

ON t o 1—* - j 4NO

O ON U) r-f-a>*-»

33a

Ml*< <2

01 t o to i—i t o to _l H 1—1%

p30n>

a*»c*«W9

3 a*S' »P* "*

O on to On o On On 4 U) a *■*» M Oto ON to ►—i 4 4 4 4 o l

3*1 O ^

° 5* ^ a

to o O to 1—1 On F 23* p

B 5.o 1—1 00 NO ON o> ■O -O NO 0 f j fD Q.o NO -o o U) NO OO O n O n 0 p Vi W> 3

ts> Vi rt- r? p»Vi

P3

P M3 «®

NO Jin t o u> -U O n 4 O n cn 3 aVi

Q . O p “*j~o t o 1—1 OO 00 —1 00 O n NO II

O

B «aNO to U) 00 On 00 00 O nO

t o P3a

p■1

J P0. sSS’ m -

pa.»9

a £ a- 89Ol NO ON NO i—i NO 00 i—i -0 00

-U - o U) -U u>4 'O U)O n

ca*fD*-i

sMfMna

^ *3"ip »

pM) A

*3 P

o o o o o O 0 O 0

0i-h 30e ^ H-o O) o l—l l—l ’—1 t o l—l K 3 3 g2 8 » S 2. B

On U> -o O n to o> t o 4 Cl*02

O o o O o O 0 to O & < ’ 2 S '4 t o i—I t o 4 O) '4 0 ON 0 n — s rNO 4 U) NO 00 ■0 U) OJ 0 3 sC+

Vi O'na.

f6 Mas c« 0 *

i— i 1— 1 1— 1 l—l i— i l—l l—l l—l 1—1 C/3 n 3 “ g ,00 O) U> NO ON O n NO 4 t o !“►

p3a.

0n■»a

fp2 3 S ’. 30 33 31—1 1— 1 l—l l—l I— i 1— 1 l—l l—l 1— 1 C/3 3* cr

U> O l ON to 4 ON 4 co '0l—t

PC/33n

0 « ? 2 s . 3- S®55*® »w2 o-5**® «« 2. f»

£ &■ "*3 c S*0 2 3S* ** j§1 13 ^^ « rt3 Ô QTQ

3<T>3CDO'>-lBSCDno33e»C/)

§oS’3

p

o ’:5~o1>5

c§

VlT 3BJOtr

Odontocerum

0 0

0.2+/-0.5 0.2+/-0.5

albicorne(Scopoli)

ST £a Cl

O V;^ 55 a 2Co ^

U)+io

OO

oON+

vo

a -■7

Orso&Cl05

0 U) +1o

OO

toOl+IOl

OO

O)I

O n

4^

O to

to

to to

~4. 1 3

O: OE ^cd ^

"O ^ ^ c 3 ^3 , ‘CO

ci'O

0 *u» +1o

to+I4^

4^

o oO) -o l—l+ +_

o l—l-o Ol

to

to

5 r"7 .© O Oq"7 rts

&S1

CD3&

S?$•©33V«Co

o o 1—1 l—lu> ■o o>

+ + o1 1o 14L. 1to 00 41* •o

f ft §O <D a* 3:

toto+Ito

OO

to

O Ol

o -o

O VO

to VO

Oll-l 4~o ,+

I

toOOtoI

U)U)

Ol vo•o to

o1UJ

1to4L.vo

o 4^to 00 o>+_ +_

1o

1Ol OO

Ol U) l—l

VO

a g 1Oa ss- w S*0

1

VOo>+

O n.+i00u>

toOll—l+ItoOl

O n

OO

+

O n

00

ON

PCl3

fflRo3

too ON VO i— i

U) ■o 00 U)+ + + +1o 1

l— l100 1to

O n 4Ô n

O n VO

U)■to

VO

UJIOl

Ol

Io4^

to

to

ocCD

&oo

OO<“+&3a .

cocO*CD

Oc

oo

cor-f33a*

ooc■-tc rCD

H

p

£3a§oCD

2oo’-V

a 2 ^ < < 2 . 5* » <g. ft » S0 G.3 p«< °S o - « . * ~ nH f f (9 © CD 3w v, «■*■3 2 33 5 s®1 a 9o- 2 3.« ™ £ . ■7 CZ> O° s? 3- ►2 CD O ' o -1 gfi «j 3=5 g a .2 3 *®CD ~ mS - «G rt »cP **" 2 s ^M H £■^ O csr • cd cd 0 »s 3>^ M* W5 3 "*3 a ^b * 5* ! •CD 5 . s*** a ©o e cd; s , « » “ £ S'S 5 73 S$ a .X. ^ M3 <« a 9989 « g r o aS 7 )2 s c| « ’ l 2c a*3 a. S g*^ 3 ‘ ► ‘S .

-3 0 “ ^ - d CD O MM 2 M G “ o CD D £" S c= g s J L a lo TO’ 5 31CD *

3 3-*7 CD CD 3

i s i3 3t« < -+ O > -»

2.

a ^3 cr•7 CD B . O.

► ^ O' »s S3 ?T°* rK fig ON 3 i2 tOCD •CD«< ^3 ^ci X3 3CD _<» 3!5 s ^ ■» oM “**g 53 < o . »% 3-5 CD2- o'ff. *7 O 35 CD

M3

&* A 2 CD 2 O2 B3 2 3 -•CL 3to TO•7a 5*CMQ- »S ’ ^^ §; CD S'*

TO* ^b * a^ aO Qtq"*» a1- fe3 M,

^ 2 3 § R a .CD 2 . » SL s

a -B oCD 3 3r> c 3 “ 3M3 ^CDCD ®9CD*-• P4-OB if

9 ao• 30 3“

2S* “* 3 36

a ee 52- 3*M3 rD

■a **“7 3TO 31 S’CD (TJ

3<n>3o '

oo2333

V)2nr"+3►-to

?

era

733ndrn>

^ 3“ O^ O Si c ftr

« I I-i 55 *5^ s .

p a §CL S ’0n

O l+

to

155ft->!

Oatft?s

£c£ 3

1*Cn’

15r0 3

pS3N

Of t 4 3

aPS3N

<Oft* 3

3On’

&|? r

<*•■4 .ft3

ftj55ft-ftXoT-n ft

55 53 CD*l-l SiftcT*—1 Sift

CDft 55

OnCD►1 ft

55On

SS3

1 **4 .ftr

ssOn 1?

p “ SiS 'Vi

siCn p

CL3<55a

p * ftftSiPCL P *•-»

*•*4. 0 0*p

Zn w

O O O OU ) to to to+ + + +1

O1

O1

01

0O n O n 4 L ON

to4io-JL

+

'O

to

4l+io

NO

oOO+

O O OO l - O t o 4 L+ 4 ^ + + _1

t—»■

I—*1

01

O lt—* ON O l

4lNO4IO n

O 0 0 l—lLk) ’l—l I— to+ 4^ +0 O 0

• 0 4L 4L - 0

to

to

O OJ

to to OO

to to 0 0

0 u> +1001

0 to +1o-fL

O n

Iovo

4L4-

to

to

Ol+Io00

004 -

001+

0U )+1obo

OO4 -

O n

to to + ° I

to

JLto4 “

U )

NO

ON

NO t o

NO O

4L4 -

to+

NO

.+I

U )

ON

ft ft- Cl r f t►ft On O55 ft ft- 0ft O & ftft. ft ftOsiOn

1?

On *•4.OC3

SJ

03ft

ft-*•*4.s -C/3"

0 O O 0O l NO O l u >+ + + 0 4 :1

I— *1 1

l— l1

0

O n N O U ) O l

0 to 4L 0

to U) 00 l— l

+ +0 4 l1

01to 1<1 1

0

4L • 0 0 1 u>

to 0 0

Ol 1—1 »—> I_1+ + + +1to 1 ■

01

0

Ol Ol U) U)

O l ON O OO n 4L 4L+ 4 ; + +14L 1

NO■O 1

0

O l 'O U ) Ci

1— * 0 l— l l— l

I— * O l ~ o to+ 4- 4 - 4 -

to

to

o00

to

to

to

O n - O O

to ON O i t—>

-(L 00 OO

-JL - o 4l O l ON

Invertebrate com

munity

structure in

Callitriche stagnalis

patches

Taxa Abundance

No. of

occurrences

OQOOc rpa><-+■P

ffi2CL5 'pa>

O

►3’ooBCLPn>

C/33c

CLPOf

£2

i

oa

S'a

yVSa ?aaro

£3a

"Ph ^ w a_ C5^ S 'O Sie*o a 5LjCNS3

Q. 9 o g 5. a?& « 5* 38 B * ^ |p s* a

x— OjpCL

a 2n> a

&!oS 'Co

w g .P Kft- CL ^ 3

£bSTCo

o l—l t o ou> t o 4L OJ vo l- l+ 1—* 4L O l 1 +

11 l—l 00 l—l 'Oo0 0

VO +i

JL +1

oo j

to4L+UJt o

+_io

Os)

U)+

00

-P t oO l JL U) 4L00 Ol Ol U)vo + Ol +

0 t o +1o4L

0 t o

1oJL

+Io

OJ

Ol+I

o \o

Ol00

VO

VO 0 0

Os) Os) i—i

+

1

+

1

0 0

+1—1 1-^

o I—* 4LOs) O n

00 + I

01 4L

-0+Io

vo

Ov+

oU)

t o+Io4L

Ol+

vo+

00

obo+

l—l Os)to vo Os) l—l'O 'O oO ON o 004L + -o +

t o+Io4L

Ol+I

O l4L

o p1—IOl

- o l—l -pL+ +

1o

1o

1l—l

vo Os) 4LOl

t o+Io

4L

Ol+

Ol t ol—lON 00

t o ON OJ +1+ + +1 1 1 l—lt o o t o Os)- o - o l—l vo

Ol

l—l o o Ol ovo I—I to OO l—l

+ + +

to o o Ol o'l—l 4L - o Os) 4L

0 t o+_1001

O l 4L o v •—1 •—1■ O 00 o

00 VO 4L O) O vo 4L O ' O O

OO VO 4L V O -JL

o o -pL - o O O n t o

Invertebrate

comm

unity structure

in Callitriche

stagnalis patches

Taxa Abundance

No. of

occurrences


6.3.3 Community structure

Samples taken from the outer, mid and root sections of the stands and surber samples

were separated by Detrended Correspondence Analysis (DCA) (Jongman et al. 1987)

using their macroinvertebrate communities, see Figure 6-3. Root section and Surber

samples were not separated by the DCA, but 3 species did show a preference for the

Surber samples over the root sections, see Table 6-3.

Table 6-3. Taxa with a significant preference for Surber samples over root sections (Wilcoxon matched pairs test, N = 10).

Taxon z Pvalue

Baetis rhodani 2.80 0.01Rhyacophila 2.20 0.05Limniusvolckmari

2.20 0.05

_ i_ Min-Max 1 I 25%-75%

OUTER MID ROOT D Median value

Figure 6-1 Number of invertebrate taxa no./g fresh weight. Outer was significantly different from root, Wilcoxon Matched-Pairs test: p<0.05, n=10. Values standardised for section weight.

0.14

CO 0.10XCOh-

N—oQ 0 . 0 0

0.02

173


H I Min-Max d j 25%-75%

OUTER MID ROOT D Median value

Figure 6-2 Relative abundance of invertebrates no./g fresh weight. Outer was significantly different from mid and root, Wilcoxon Matched-Pairs test: p<0.05, n=10. Values standardised for section weight.

To standardise between Surber and stand samples, sample area was used, but section

weight was not taken into account. Other analyse showed that the same separation

occurred if section weight was also used but separation of sites was not so clear.

Samples from the outer section were most dissimilar to the Surber sample, the root

section was most similar. The outer section was dominated by the filter feeding

Simuliidae while in all other plant sections Ephemerella ignita dominates. In the

Surber samples Baetis rhodani was the most abundant invertebrate. Community

structure also differed between sections, see Figure 6-4. All three stand sections

supported invertebrate communities which followed the broken stick model (Gray

1987). The outer section, which is exposed to the highest velocities, was the least

equitable with the steepest gradient at the start and is the closest to a geometric series.

Such a series is found in communities poor in species at the earliest successional stage

after colonisation or when exposed to extreme conditions (Gray 1987). The middle

section is more equitable and the root section is the most equitable having the largest

174


lDCO+

elmad

+

ephign

f ■ •_ O

gampul

sim + + + + +leutus

♦o

aseaqu

chi

limvad

oultul i

°0bra°^ o

lim / tip

ecd

baerho

hydsil

• . \ | oHhir nmvla

♦ ♦ath esopar

caenv ancflurhy

psi

lOO J globol

- 0 . 5 + 4 . 0

Figure 6-3 Detrended Correspondence Analysis Site by species biplot, (eigenvalues: axis 1 = 0.6, axis 2 = 0.136) Plant section and Surber symbols: Outer +, Mid ■ , Root 0 and Surber O. The location of each taxa is marked with ♦ and a taxa code. Taxa codes: Caenis rivulorum (caeriv), Baetis rhodani (baerho), Ephemerella ignita (ephign), Ecdyonurus (ecd), Hydropsyche siltalai (hydsil), Athripsodes (ath), Brachycentrus subnubilus (brasub), Rhyacophila sp. (rhy), Glossosoma boltoni (globol), Asellus aquatic us (aseaqu), Gammarus pulex (gampul), Ancylus fluviatilis (ancflu), Pisidium sp. (psi), Limnius volckmari larva (limvla), Limnius volckmari adult (limvad), Elmis aenea larva (elmaen), Elmis aenea adult (elmad), Oulimnius tuberculatus adult (oultuad), Oulimnius tuberculatus larva (oultula), Esolus parat/e/pipedus (esopar), Leuctra fusca group (leufus), Tipulinae (tip), Limoniinae (lim), Simulidae (sim), Chironomidae (chi), Hirudidae (hir), Oligochaeta (oli).

175


number of species which share the resource more evenly, in terms of producing

individuals.

6.3.4 Prevalent flow conditions

It is suggested that stand structure and velocity combine to create a range of stability

conditions, forming a series o f microhabitats, in, on and outwith the plant beds.

Velocity was significantly different and noted to be lower over the top of the stands

than directly in front o f the stands or in the adjacent Surber samples (Wilcoxon

Matched Pairs p<0.05, n = 10). It is suggested that the shape of the stand is deflecting

water to either side and creating flow conditions which are similar to sheet flow over

the top of the stand; that is turbulent but uniform in direction o f flow. The velocity at

the top of the stand (0.25 ms'1) was higher and significantly different (Wilcoxon

Matched Pairs p<0.05, n= 6) than that in the middle (0.046 ms'1) but velocity at the

bottom (0.040m/s) of the stand was not significantly different from that in the middle.

The near bed velocities at the Surber sample points (0.042 ms'1), (estimated by

adjusting velocity measured at 0 .4 of depth from the bed) are not significantly different

to those found in the mid and root sections of the stand. Estimates o f near bed velocity

over substrate were an order of magnitude lower than those measured on the outside

of the plant stands. An estimate of the required velocity at 0.4 of depth above

substrate flowing over unvegetated substrate that would produce a near bed velocity

similar to the velocity experienced by the outer section of the stands would equate to

2.5ms'1, or near spate condition, see Figure 6-5. This figure was derived from using

Carling's (1992) estimate procedure in reverse starting with the flow over the outer

section of the stands (0.25 m/s).


0.1t=OJ73

CDa:

0.01

0.001

ono o c,

n o , .) o o lJ a a I

:: o o 6 ... :Q O

12 16 20

Rank

24 28

O OUTER

= MID

ROOT

Figure 6-4 Rank abundance plot for the 3 plant sections. Relative abundance of invertebrates no/cm2/g fresh weight.

0.9

co U.uEuo3 0.3 >

~T~ Min-Max

! i 25%-75%

a Median valueSURBER FRONT OUTER MID ROOT

Figure 6-5 Velocity m s1, 50 second means.

6.4 Discussion

The presence of Callitriche stagnalis stands in the Blane Water increases the diversity

of invertebrate habitat within the river. Taxa which occurred in the plant stands also

occurred on the substrate, but some were present in higher numbers, on and within

macrophyte beds. The stands extend the available substrate for invertebrate

colonisation up through the water column, providing a continuum of conditions from

177


high to low velocity. Adjacent benthic substrate does not provide the high velocities

present on the outside of the plant stand in a sustained manner, that is only under near

spate conditions would these high velocities be achieved. In addition the external

surface of the stands greatly increases the area of high-flow substrate available for

colonisation by invertebrates adapted to high flow conditions, especially Simuliidae

(Hart & Latta 1986). Their habitat requirements include the need for fast flowing

water through their labral fans but they avoid areas o f maximum shear stress (Craig &

Chance 1981; Lacoursiere 1991) and it is likely that the correct complex of hydraulic

conditions are available on the surface of the plants.

Although taxon richness and abundance were not significantly different between the

mid and root sections o f the plants they were separable using DCA. The rank

abundance plots demonstrated that the root section was more equitable than the mid

section in resource distribution, hence the separation in the DCA biplot. This would

suggest that conditions were more extreme in the mid section than the root section,

which could be a product o f longer term conditions than those measured in this study.

The conditions present in these two sections allowed E.ignita to dominate where B.

rhodani dominated in the Surber samples.

It has been suggested by (Sand-Jensen & Madsen 1992) that Callitriche cophocarpa

"patches form a mutually protecting structure against high flows“ and “may,

therefore... increase nutrient supply from sedimentation fine-grained particles". He

later found that the sediment under stands often has high levels of organic matter

(Sand-Jensen 1998) which could provide food for detritivores. In this study, despite

finding that the substrate under our stands was more fine grained than that in

unvegetated areas we found that no detrivore showed a preference for the root


section, compared to the adjacent unvegetated sediment. I did not measure the amount

of organic matter in the sediment.

The net result o f the presence o f the plants is to increase the invertebrate abundance

and diversity present in the Blane Water. The assemblages present can be related to

the structure o f the stands, and how this influences the flow conditions experienced by

the animals. The complexity of invertebrate community supported appears to be

functionally related to position within the stand, with extreme conditions (on the

external surface) supporting fewer species than within the bed. The total diversity

supported is however a function of the total available set o f conditions present in the

river: on, in and between the stands.

From a management point o f view it is clear that the presence of the stands

substantially increases the invertebrate diversity and abundance supported by the river,

particularly by increasing the habitat available for organisms (e.g. Simuliidae) which

would otherwise find only a limited space for colonisation in their preferred high-flow

conditions.

Callitriche stagnalis is a common species in British rivers, and elsewhere in Europe,

with a high tolerance of disturbance (Sabbatini & Murphy 1996). Inclusion o f

transplanted material as part of rehabilitation schemes in the small gravel-bed rivers

which are well-suited to this species would be likely to make a useful contribution to

improving invertebrate support function of the river.

6.5 Conclusions

• The macrophyte stands supported a greater abundance of benthic macroinvertebrates than bare substrate.

• Higher velocities were recorded on the surface of the macrophyte stands than on the bare substrate.

179


• The invertebrate community occurring on the outside of the stands was dominated by Simuliidae and was less equitable than the community living on the interface between the plant stands and the substrate.

• By diversifying the habitat for benthic invertebrates, macrophyte stands are a potentially useful tool in river rehabilitation.

180

General Discussion

Chapter 7: General Discussion

7.1 Review of results

The original aim of this study was to provide data on flow preferences of benthic

invertebrates which could contribute both to our ecological understanding of these

organisms, and assist in construction of river management tools. In Chapter 2 ,1 attempted

to define deep and shallow river sections as riffles, pools and runs, using a range of

criteria, and came to the conclusion that visual observation, coupled with depth and

velocity measurements was necessary. Despite the minimal differences in benthic

invertebrate community between the deep and shallow reaches of rivers used in this study

it is likely that, indirectly at least, deep and shallow reaches do form important habitat

units for the benthos. Predatory fish, salmonids, have a direct effect on the benthic

community by catching animals drifting out of riffles into pools (Ade 1989); they also use

riffle areas as spawning grounds. It has also been suggested that riffles provide more

refugia and are less disturbed than pools (Scarsbrook & Townsend 1993).

In Chapter 3, I demonstrated both the relative homogeneity of the invertebrate community

within each river, and also the existence of gradients between erosional and depositional

conditions, with taxa occurring along these gradients: a situation found in similar studies in

Australia (Barmuta 1989). The results of the analysis presented in Chapter 4 suggest that

data collected using a randomised sampling strategy on instream habitat is not suitable for

precise quantification of invertebrate flow preferences. The methods are sufficient to

identify habitat preferences in a general manner - whether the animal prefers fast or slow

conditions - but not quantitatively. This type of study does inform us of the relative

importance of the environmental variables and how they interact with one another.

General Discussion

Laboratory flume experiments as carried out in Chapter 5 were also necessary to confirm

findings made in the field. The approach taken in Chapter 5 allows for some measure of

the plasticity in species responses to adverse flow conditions but does not offer a velocity

gradient to the animals (which would be the most useful approach). The results also

showed that the taxa had very different tolerance ranges for velocity and that behavioural

control of drift may be possible. The final results chapter (Chapter 6) looked at a natural

gradient of flow conditions in an aquatic macrophyte stand. The results presented in this

chapter suggested that the higher velocities found on the outside of plant stands are not

preferred by some species but were by others, e.g. Simulium. Whether this was due to

velocity or some other another factor such as the presence of predators is difficult to

elucidate from the present study and requires more work. Exposing the animals in a flume

to a velocity gradient which covers the range found in the macrophyte patch under

controlled conditions would go some way to testing the theory that water velocity

preference is one of the main factors influencing benthic invertebrate occurrence.

7.2 Future Work

Identifying flow preferences of benthic invertebrates remains as a major aim necessary for

the ecology and management of lotic freshwater systems. A refined version of the

approach taken in this thesis could supply the information required.

A potential approach to elucidating the instream distribution of invertebrates involves both

increasing and decreasing the spatial scale examined. To demonstrate clearer differences in

invertebrate community structure one would need to examine a wider range of conditions

e.g. in the rivers used in my study the deepest pool section was 1.5 m but a deeper range

would be more informative, see Chapter 2, section 2.3. Large scale studies as suggested

182

General Discussion

here have previously identified habitat units at the start of the study and based their

sampling round these; edge, riffle, pool etc. A greater emphasis on covering wide ranges

of parameters of interest, such as velocity, depth, substrate type and periphyton cover is an

alternative approach advocated here. Reducing the scale of sample area and thereby

making velocity measurements more accurate is also necessary and would seem sensible.

Which spatial scale is correct for this type of procedure could be determined by a nested

sampling design where the spatial scale is reduced at a number of levels. This type of study

has been done at large physical scale, between rivers and reaches but not for instream

habitat. Elliot (1977) gives a number of methods for determining whether benthic

invertebrate distribution is aggregated or not and these could be applied to such data. It is

likely that different taxa will exhibit clumping at different spatial scales and that a single

taxon can also clump at different scales, e.g. in Chapter 2.3.6, p47, B. rhodani showed a

preference for run sections in the Duneaton Water and Blane Water which is possibly

caused by the presence of suitable, smaller habitat units: rocks with periphyton, which if

examined at this scale the animals would also show a preference. This type o f study could

potentially use some of the new sampling methodology available such as stereo

photography which allow fine scale measures of invertebrate distribution (Evans & Norris

1997). Direct measures of turbulence were not included in this study and it is proposed

that any future work would include small scale localised measures of turbulence, which

given the potential of turbulent water to surprise invertebrates and entrain them must mean

areas of high turbulence are avoided. By turbulence here I do not simply mean non-laminar

flow but gross, random motion of water at scales larger than the molecular.

183

General Discussion

By concentrating on individual species and recording their velocity and depth tolerances

under controlled laboratory conditions we could begin to determine the importance of

biotic and abiotic factors in their distribution. The lack o f this type of basic knowledge

hampers our interpretation of large survey work as carried out in this study. Based in the

laboratory the work could proceed faster than trying to make accurate measurements of

single species distributions in relation to velocity in the field, as suggested in the previous

paragraph, which requires painstaking and time consuming methods, e.g. Hart et al.

(1996). Field based measurements can only be applied to animals while they are on the

surface o f the substrate and, as many species spend large portions o f time among the

substrate matrix, a second series of experiments would be necessary to identify the

proportion of time spent in both locations.

The spatial complexity o f instream habitat has been linked to invertebrates’ ability to

withstand disturbance events (Lancaster & Hildrew 1993b). It is possible that streams of

different retention capacity support invertebrates with different life history strategies. The

homogeneity of instream habitat in channelled rivers could therefore cause the strategies

of these organisms to be altered and could exert a selective pressure on the animals. The

implications this has for the genetic variability or life history strategies of these animals has

yet to be addressed. New work is required to look at this question which must incorporate

studies o f both juveniles and adults.

7.3 Management recommendations

This study aimed at a very specific question of river management: how to improve

instream habitat for benthic macroinvertebrates. There are three major impacts on this

184

General Discussion

habitat type in small Scottish streams; river channelisation and alteration of channel

discharge either by damming, changes in catchment land use or water abstraction, often

for fish farms. Management solutions applied in Scotland include the maintenance of

minimum discharges (Q90 and Q95) from reservoirs and in streams from which water is

abstracted, as imposed by SEP A (Marsden, M., SEP A pers comm) and the physical

rehabilitation of channelised systems (Gilvear & Bradley 1997).

To date management guidelines suggested by public bodies (SNH and SEP A) and Non

Governmental Bodies (e.g. RSPB) have mainly concentrated on physical habitat

improvement for taxa other than macroinvertebrates, usually fish, birds and vegetation

(Hoey et al. 1998). With the forthcoming EU Water Framework Directive which

emphasises the ecological and physical naturalness of surface waters there will be a

growing need to improve instream rehabilitation for invertebrates.

The importance of aquatic macrophytes as a useful tool in the conservation of

macroinvertebrates has been previously highlighted (Wright et al. 1995). The results of

Chapter 6 emphases this fact and illustrate the need for more basic research in this area if

we are to manage aquatic vegetation for the benefit o f macroinvertebrates without

compromising the capacity of channels to deal with high flows; a problem which has a

number of potential solutions (Fox & Murphy 1990a; Fox & Murphy 1990b). The density

of macrophytes can even be incorporated into PHABSIM models if one wished (Hearne et

al. 1994)! If used correctly macrophytes could potential mitigate the impacts of

disturbance caused by rising discharge in a similar manner to woody debris (Borchardt

1993) and potentially increase the number of macrophyte species too (Hey et al. 1994).

185

General Discussion

Structural rehabilitation schemes are usually very successful in improving the diversity and

productivity o f the altered reaches (Swales & O'Hara 1980). The process of community

alteration resulting from structural alterations to channels as part of habitat enhancement

schemes can be difficult to judge in detail (Lynch & Murray 1994). If we wish to advance

the science of this process and its usefulness as a management tool studies such a those

advocated in section 7.2 will be necessary to improve their design.

PHABSIM is one of the few models which is now used to predict the amount of useable

habitat for invertebrates at different discharges. In Chapters 3 and 4 I identified a number

of problems with the data on which this model is based which need to be addressed if the

model is to more accurately predict useable habitat for invertebrates. The previous section

on further research suggests the type of projects necessary to collect this data.

There is a more fundamental problem associated with using this type of model and that is

that its aim is only to make sure there is suitable habitat available for the animals. It does

not take into account the dynamic nature of lotic systems which fuel the natural evolution

of the invertebrate community e.g. it is quite possible for the same Q90 to be realised daily

without variation. As ecologists our advice to managers should be to keep the physical

system as natural as possible and the invertebrates will adapt to it. For the present it may

be more suitable to use models based on historical flows or the flows implied by the rivers

geomorphology, (see Jowett (1997) for a review of approaches). Geomorphologists have

a more developed understanding of how any particular river should look and act than

ecologists have of how animals interact with the habitat provided by the river (Werritty et

al. 1994). Given the present state of knowledge in both disciplines ecologists should

186

General Discussion

support the work o f managers where they try and recreate the natural geomorphology of

rivers.

Frequently this approach is too simplistic as it can be impossible to recreate a river’s

‘natural’ geomorphology given the small sections of rivers that can be worked on and the

numerous impacts presently outwith the control of managers, e.g. climate change and

catchment land use (Langan et al. 1997). There is also the possibility that by increasing

flow variability one could decrease available benthic invertebrate habitat (Englund &

Malmquist 1996). This emphasises the need for an improved understanding of the

interaction between benthic invertebrates and the physical habitat of rivers is necessary.

The approach taken in this thesis utilised a direct assessment of the habitat usage by

invertebrate organisms: to develop suitable management models indirect interactions also

need to be understood. The key to this understanding is the role of natural disturbance

(flood events), interactions with riparian ecotones and instream habitat patchiness. Recent

research which has started to describe the mechanisms underpinning these interactions was

discussed at the start o f this thesis (Chapter 1, section 1.1): it appears that much more

work is required in this field.

Recent papers on the state of freshwater ecology and the management of freshwaters

emphasises the need for ecologists to champion the relevance of their work and to review

the paradigms of aquatic ecology (Reynolds 1998). This work also emphasises the need

for 'quantitative bench marks and empiricised strategies to be set'. Reynolds (1998), similar

to the approach take in the restoration of the Danube fish populations, (Keckeis &

Schiemer 1992; Schiemer & Wieser 1992; Tockner & Schiemer 1997). This promotion of

systems ecology has been heeded by UNESCO which now has a work program which is

187

General Discussion

developing the tools and the paradigm of 'Ecohydrology'. Among the many UNESCO

projects are those that look at the interactions of river flow instream and in the wider

context of the catchment.

In summary the advice given to managers of aquatic ecosystems is not new and is that

more basic scientific research is needed (Welcomme 1992). That UNESCO considers an

ecologically balanced and soft engineering approaches to river management o f importance

emphasises the need for both more research and sensitive management.

188

References

References

Adams, C. E. & Maitland, P. S. 1998 The ruffe population of Loch Lomond, Scotland: Its introduction, population expansion, and interaction with native species. Journal o f Great Lakes Research 24, 249-262.

Ade, R. 1989 The Trout and Salmon Handbook. A Guide to the Wild Fish. Bromley: Christopher Helm Ltd.

Ambuhl, H. 1959 Die Bedeutung der Stromung als okologischer Faktor. Schweiz Z. Hydrol 21, 133-264.

Armitage, P. D. 1976 A quantitative study of the invertebrates of the River Tees below Cow Green Reservoir. Freshwater Biology 6, 229-240.

Armitage, P. D. 1995 Faunal community change in response to flow manipulation. In The EcologicalBasis o f River Management (ed. D. M. Harper & A. J. D. Ferguson), pp. 59-78. Chichester: John Wiley & Sons.

Armitage, P. D., Cranston, P. S. & Pinder, L. V. C. 1995 The Chironomidae : Biology and Ecology o f Non-biting Midges. London: Chapman and Hall.

Armitage, P. D. & Gunn, R. J. M. 1996 Differential response of benthos to natural and anthropogenicdisturbances in 3 lowland streams. Internationale Revue Der Gesamten Hydrobiologie 81, 161- 181.

Armour, C. L. & Taylor, J. G. 1991 Evaluation of the instream flow incremental methodology by United- States fish and wildlife service field users. Fisheries 16, 36-43.

Bacher, I. & Waringer, J. A. 1996 Hydraulic microdistribution of cased caddis larvae in an Austrian mountain brook. Internationale Revue Der Gesamten Hydrobiologie 81, 541-554.

Badcock, R. M. 1976 The distribution of the Hydropsychidae in Great Britain. In Proceedings o f the First International symposium on Trichoptera (ed. H. Malicky). The Hague.

Barmuta, L. A. 1989 Habitat patchiness and macrobenthic community structure in an upland stream in temperate Victoria, Australia. Freshwater Biology 21, 223-236.

Begon, M., Harper, J. L. & Townsend, C. R. 1996 Ecology: Individuals, Populations and Communities. Oxford: Blackwell Science.

Biron, P. M., Lane, S. N., Roy, A. G., Bradbrook, K. F. & Richards, K. S. 1998 Sensitivity of bed shear stress estimated from vertical velocity profiles: The problem of sampling resolution. Earth Surface Processes and Landforms 23, 133-139.

Borchardt, D. 1993 Effects of flow and refugia on drift loss of benthic macroinvertebrates - Implications for habitat restoration in lowland streams. Freshwater Biology 29, 221-227.

Borchardt, D. & Statzner, B. 1990 Ecological impact of urban stormwater runoff studied in experimental flumes - population loss by drift and availability of refugial space. Aquatic Sciences 52. 299-314.

Bouckaert, F. W. & Davis, J. 1998 Microflow regimes and the distribution of macroinvertebrates around stream boulders. Freshwater Biology 40, 77-86.

189

References

Boulton, A. J., Peterson, C. G., Grimm, N. B. & Fisher, S. G. 1992 Stability of an aquaticmacroinvertebrate community in a multiyear hydrologic disturbance regime. Ecology 73, 2192- 2207.

Boumaud, M. 1963 Le courant, facteur ecologique et ethologique de la vie aquatique. Hydrobiologia 21, 125-65.

Bovee, K. D. 1982 A guide to stream habitat analysis using instream flow incremental methodology.United States Fish and Wildlife Service, Cooperative Instream Flow Group, Instream Flow information paper 12, 248.

Briggs, J. C. 1948 The quantitative effects of a dam upon bottom fauna of a small Californian stream.Transactions o f the American Fisheries Society 78, 70-81.

Brinkhurst, R. O. 1963 A Guide fo r the Identification o f British Aquatic Oligochaeta. Scientific Publication No. 22: Freshwater Biological Association.

Brittain, J. E. & Eikeland, T. J. 1988 Invertebrate Drift - A Review. Hydrobiologia 166, 77-93.

Brooks, S. S. & Boulton, A. J. 1991 Recolonization dynamics of benthic macroinvertebrates after artificial and natural disturbances in an Australian temporary stream. Australian Journal o f Marine and Freshwater Research 42, 295-308.

Brown, A. V. & Brussock, P. P. 1991 Comparisons of Benthic Invertebrates Between Riffles and Pools.Hydrobiologia 220, 99-108.

Brown, H. P. 1987 Biology of riffle beetles. Annual Review o f Entomology 32, 253-273.

Brusca, R. C. & Brusca, G. J. 1990 Invertebrates. Sunderland, Massachusetts: Sinauer Associates, Inc.

Buchan, L. A. J. & Padilla, D. K. 1999 Estimating the probability of long-distance overland dispersal of invading aquatic species. Ecological Applications 9, 254-265.

Buffagni, A., Crosa, G. & Marchetti, R. 1995 Size-related shifts in the physical habitat of 2 mayfly species (Ephemeroptera). Freshwater Biology 34, 297-302.

Carling. 1992 The nature of the fluid-boundary layer and the selection of parameters for benthic ecology. Freshwater Biology 28, 273-284.

Carling, P. 1995 Implications of sediment transport for instream flow modelling of aquatic habitat. In The ecological basis fo r river management (ed. D. Harper & A. J. D. Ferguson). Chichester: John Wiley & Sons.

Carling, P. A. 1993 The nature of the fluid boundary layer and the selection of parameters for benthic ecology: Reply to Lamouroux. Freshwater Biology 30, 333-335.

Ciborowski, J. J. H. 1983 Influence of current velocity, density, and detritus on drift of 2 mayfly species (Ephemeroptera). Canadian Journal o f Zoology 61, 119-125.

Ciborowski, J. J. H. 1987 Dynamics of drift and microdistribution of 2 mayfly populations - a predictive model. Canadian Journal o f Fisheries and Aquatic Sciences 44, 832-845.

190

References

Clausen, B. & Biggs, B. J. F. 1997 Relationships between benthic biota and hydrological indices in New Zealand streams. Freshwater Biology 38, 327-342.

Clegg, J. 1952 The Freshwater Life o f the British Isles. London and New York: Fredrick Wame &Co.,Ltd.

Clifford, N. J. 1993 Formation of riffle pool sequences - field evidence for an autogenetic process. Sedimentary Geology 85, 39-51.

Cobb, D. G., Galloway, T. D. & Flannagan, J. F. 1992 Effects of discharge and substrate stability ondensity and species composition of stream insects. Canadian Journal o f Fisheries and Aquatic Sciences 49, 1788-1795.

Collier, K. J., Croker, G. F., Hickey, C. W., Quinn, J. M. & Smith, B. S. 1995 Effects of hydraulicconditions and larval size on the microdistribution of Hydrobiosidae (Trichoptera) in two New Zealand rivers. New Zealand Journal o f Marine and Freshwater Research 29, 439-451.

Collier, K. J. & Wakelin, M. D. 1992 Drift of aquatic macroinvertebrate larvae in Manganuiateao River, Central North Island, New Zealand. New Zealand Natural Sciences 19, 15-26.

Cowan, C. A. & Peckarsky, B. L. 1990 Feeding by a lotic mayfly grazer as quantified by gut fluorescence. Journal o f the North American Benthological Society 9, 368-378.

Craig, D. A. & Chance, M. M. 1981 Filter feeding in larvae of Simuliidae (Diptera: Culicomorpha):aspects of functional morphology and hydrodynamics. Canadian Journal o f Zoology 60, 712-724.

Crawley, M. J. 1986 Plant Ecology. Oxford: Blackwell Scientific Publications.

Crowl, T. A. & Schnell, G. D. 1990 Factors determining population-density and size distribution of a fresh-water snail in streams - effects of spatial scale. Oikos 59, 359-367.

Crowl, T. A., Townsend, C. R.. Bouwes, N. & Thomas, H. 1997 Scales and causes of patchiness in stream invertebrate assemblages: Top-down predator effects? Journal o f the North American Benthological Society 16, 277-285.

Cudney, M. D. & Wallace, J. B. 1980 Life cycles, microdistribution and production dynamics of sixspecies of net-spinning caddisflies in a large southeastern stream. Holartic Ecology 3, 169-182.

Cummins, K. W. 1973 Trophic relations of aquatic insects. Annual review o f entomology 18, 183-206.

Dahl, J. & Greenberg, L. 1996 Effects of habitat structure on habitat use by Gammarus pulex in artificial streams. Freshwater Biology 36, 487-495.

Davis, J. A. & Barmuta, L. A. 1989 An ecologically useful classification of mean and near -bed flows in streams and rivers. Freshwater Biology 21, 271-278.

Death, R. G. 1996 The effect of patch disturbance on stream invertebrate community structure - the influence of disturbance history. Oecologia 108, 567-576.

Death, R. G. & Winterboum, M. J. 1995 Diversity patterns in stream benthic invertebrate communities - the influence of habitat stability. Ecology 76, 1446-1460.

191

References

Degani, G., Herbst, G. N., Ortal, R., Bromley, H. J., Levanon, D., Netzer, Y., Harari, N. & Glazman, H. 1993 Relationship between current velocity, depth and the invertebrate community in a stable river system. Hydrobiologia 263, 163-172.

Dejalon, D. G. & Sanchez, P. 1994 Downstream effects of a new hydropower impoundment on macrophyte, macroinvertebrate and fish communities. Regulated Rivers: Research & Management 9, 253-261.

Dick, J. T. A. 1992 The nature and implications of differential predation between Gammarus-pulex and G duebeni-celticus (Crustacea, Amphipoda). Journal o f Zoology 227, 171-183.

Dick, J. T. A., Faloon, S. E. & Elwood, R. W. 1998 Active brood care in an amphipod: influences of embryonic development, temperature and oxygen. Animal Behaviour 56, 663-672.

Dick, J. T. A., Irvine, D. E. & Elwood, R. W. 1990 Differential predation by males on molted females may explain the competitive displacement of Gammarus-duebeni by G-pulex (Amphipoda). Behavioral Ecology and Sociobiology 26, 41-45.

Dick, J. T. A., Nelson, N. & Bishop, J. D. D. 1997 Introduction experiments with Gammarus spp.(Crustacea: Amphipoda) in the Isle of Man (British Isles), 1949-1995. Journal o f Zoology 242, 209-216.

Dingman, S. L. 1984 Fluvial Hydrology. New York: Freeman and Company.

Dittmar, H. 1955 Ein Sauerlandbach. Untersuchungen an einem Wiesen-Mittel-gebirgsbach. Arch. Hydrobiol. 50, 305-552.

Dodds, G. S. & Hisaw, F. L. 1924 Ecological studies of aquatic insects. I. Adaptation of mayfly nymphs to swift streams. Ecology 5, 137-148.

Dodds, G. S. & Hisaw, F. L. 1925 Ecological studies of aquatic insects. III. Adaptation of caddisfly larvae to swift streams. Ecology 6, 123-137.

Dorier, A. & Vaillant, F. 1955 Sur le facteur vitesse du courant. Verh. int. Verein. theor. angew. Limnol. 12, 593-597.

Doughty, C. R. & Maitland, P. S. 1994 The ecology of the River Endrick: present status and changes since 1960. Hydrobiologia 290, 131-151.

Downes, B. J. & Jordan, J. 1993 Effects of stone topography on abundance of net-building caddisfly larvae and arthropod diversity in an upland stream. Hydrobiologia 252, 163-174.

Downes, B. J., Lake, P. S. & Schreiber, E. S. G. 1995 Habitat structure and invertebrate assemblages on stream stones: A multivariate view from the riffle. Australian Journal o f Ecology 20, 502-514.

Dudgeon, D. 1991 An experimental-study of abiotic disturbance effects on community structure and function in a tropical stream. Archiv Fur Hydrobiologie 122, 403-420.

Dudgeon, D. 1993 The effects of spate-induced disturbance, predation and environmental complexity on macroinvertebrates in a tropical stream. Freshwater Biology 30, 189-197.

Dudgeon, D. & Chan, I. K. K. 1992 An experimental-study of the influence of periphytic algae on invertebrate abundance in a Hong-Kong stream. Freshwater Biology 27, 53-63.

192

References

Ebrahimnezhad, M. & Harper, D. M. 1997 The biological effectiveness of artificial riffles in river rehabilitation. Aquatic Conservation-Marine and Freshwater Ecosystems 7, 187-197.

Egglishaw, H. J. & Mackay, D. W. 1966 A survey of the bottom fauna of streams in the Scottishhighlands. Pt. III. Seasonal changes in the fauna of three streams. Hydrobiologia 30, 305-334.

Egglishaw, H. J. & Morgan, N. C. 1965 A survey of the bottom fauna of streams in the Scottishhighlands. Pt. II. The relationships of the fauna to the chemical and geological conditions.Hydrobiologia 26, 173-183.

Elliot, J. M. 1977 Some Methods fo r the Statistical Analysis o f Samples o f Benthic Invertebrates. Windemere: Freshwater Biological Association.

Elliot, J. M. & Mann, K. H. 1919 A Key to the British Freshwater Leeches with Notes on their Life Cycles and Ecology. Freshwater Biological Association.

Englund, G. & Malmquist, B. 1996 Effects of flow regulation, habitat area, and isolation on themacroinvertebrate fauna of rapids in North Swedish Rivers. Regulated Rivers: Research and Managment 12, 433-445.

Equihua, M. 1990 Fuzzy clustering of ecological data. Journal o f Ecology. 78 519-534.

Evans, L. J. & Norris, R. H. 1997 Prediction of benthic macroinvertebrate composition using microhabitat characteristics derived from stereo photography. Freshwater Biology 37, 621-633.

Eyre, M. D., Carr, R., McBlane, R. P. & Foster, G. N. 1992 The effects of varying site-water duration on the distribution of water beetle assemblages, adults and larvae (Coleoptera, Haliplidae,Dytiscidae, Hydrophilidae). Archiv Fur Hydrobiologie 124, 281-291.

Eyre, M. D., Foster, G. N. & Young, A. G. 1993 Relationships between water-beetle distributions and climatic variables - a possible index for monitoring global climatic-change. Archiv Fur Hydrobiologie 127, 437-450.

Fitter, R. & Manuel, R. 1994 Lakes, Rivers and Streams o f Britain and North-West Europe. Collins Photo Guide. Hong Kong: Harper Collins Publishers.

Fjellheim, A. 1996 Distribution o f benthic invertebrates in relation to stream flow characteristics in a Norwegian river. Regulated Rivers-Research & Management 12, 263-271.

Fjellheim, A., Havardstun, J., Raddum, G. G. & Schnell, O. A. 1993 Effects of increased discharge onbenthic invertebrates in a regulated river. Regulated Rivers-Research & Management 8, 179-187.

Fox, A. M. & Murphy, K. J. 1990a The efficacy and ecological impacts of herbicide and cutting regimes on the submerged plant-communities of 4 british rivers . 1. A comparison of management efficacies. Journal o f Applied Ecology 27, 520-540.

Fox, A. M. & Murphy, K. J. 1990b The efficacy and ecological impacts of herbicide and cutting regimes on the submerged plant-communities of 4 british rivers .2. A multivariate-analysis of the effects of management regimes On macrophyte communities. Journal o f Applied Ecology 27, 541-548.

Fozzard, I.R., Doughty, C.R. & Clelland, B.E. 1994 Invertebrates. The Freshwaters o f Scotland-ANational Resource o f International Significance. (Ed. Maitland, P. S., Boon, P. J. & McLusky, D.S.) John Wiley & Sons. Chichester.

193

References

Gauch, H. G. 1982 Multivariate Analysis in Community Ecology. Cambridge: Cambridge University Press.

Georgian, T. & Thorp, J. H. 1992 Effects of microhabitat selection on feeding rates of net-spinning caddisfly Larvae. Ecology 73, 229-240.

Giller, P. S. & Malmquist, B. 1998 The Biology o f Streams and Rivers. Biology of habitats. Oxford: Oxford University Press.

Gilvear, D. & Bradley, S. 1997 Geomorphological adjustment of a newly engineered upland sinuousgravel-bed river diversion: Evan Water, Scotland. Regulated Rivers: Research and Management 13, 377-389.

Golubkov, S. M., Tiunova, T. M. & Kocharina, S. L. 1992 Dependence of the respiration rate of aquatic insects upon the oxygen concentration in running and still water. Aquatic Insects 14, 137-144.

Gore, J. A., Crawford, D. J. & Addison, D. S. 1998 An analysis of artificial riffles and enhancement of benthic community diversity by physical habitat simulation (PHABSIM) and direct observation.Regulated Rivers-Research & Management 14, 69-77.

Gray, J. S. 1987 Species-abundance patterns. In Organisation o f Communities; Past and Present (ed. J. H.R. Gee & P. S. Giller), pp. 53-68. Oxford: Blackwell Scientific Publications.

Gregg, W. W. & Rose , F. L. 1982 The effects of aquatic macrophytes on the stream microenvironment.Aquatic Botany 14, 309-324.

Guinand, B., Tachet, H. & Roux, C. 1994 Longitudinal distribution and ecophysiological charcteristics of Hydropsyche exocellata (Trichoptera: Hydropsychidae) in large rivers. Ecography 17, 189-197.

Gurtz, M. E. & Wallace, J. B. 1984 Substrate-mediated response of stream invertebrates to disturbance.Ecology 65, 1556-1569.

Hammond, C. & Merrit, R. 1983 The Dragonflies o f Great Britain and Ireland. Colchester: Harley Books, Martins, Great Horkesley, Essex C06 4AH.

Hansen, R. A., Hart, D. D. & Merz, R. A. 1991 Flow mediates predator-prey interactions between tricladflatworms and larval black flies. Oikos 60, 187-196.

Hanski, I. 1994 Metapopulation Ecology. In Spatial and Temporal Aspects o f Population Processes (ed.O. E. Rhodes Jr). Georgia: University of Georgia.

Harper, D., Ebrahimnezhad, M. & Cot, F. C. I. 1998 Artificial riffles in river rehabilitation: setting the goals and measuring the successes. Aquatic Conservation-Marine and Freshwater Ecosystems 8, 5-16.

Harrison, S. S. C. & Hildrew, A. G. 1998 Distribution dynamics of epilithic insects in a lake littoral.Archiv Fur Hydrobiologie 143, 275-293.

Hart, D. D. 1991 Feeding postures of suspension-feeding larval black flies - The conflicting demands of drag and food acquisition. Oecologia 85, 457-463.

Hart, D. D. 1992 Community organization in streams - the importance of species interactions, physical factors, and chance. Oecologia 91, 220-228.

194

References

Hart, D. D., Clark, B. D. & Jasentuliyana, A. 1996 Fine-scale field measurement o f benthic flowenvironments inhabited by stream invertebrates. Limnology and Oceanography 41, 297-308.

Hart, D. D. & Latta, S. C. 1986 Determinants of ingestion rates in filter-feeding larval blackflies (Diptera, Simuliidae). Freshwater Biology 16, 1-14.

Hawkins, C. P., Hogue, J. N., Decker, L. M. & Feminella, J. W. 1997 Channel morphology, water temperature, and assemblage structure of stream insects. Journal o f the North American Benthological Society 16, 728-749.

Heame, J., Johnson, I. & Armitage, P. 1994 Determination of ecologically acceptable flows in rivers with seasonal-changes in the density of macrophyte. Regulated Rivers-Research & Management 9, 177-184.

Hey, R. D., Heritage, G. L. & Patterson, M. 1994 Impact of flood alleviation schemes on aquatic macrophytes. Regulated Rivers: Research and Managment 9, 103-119.

Hickin, N. E. 1967 Caddis Larvae o f the British Trichoptera. London: Hutchinson & Co. Ltd.

Hildrew, A. G. & Giller, P. S. 1994 Patchiness, species interactions and disturbance in the streambenthos. In Aquatic Ecology, Scale, Pattern and Process (ed. P. S. Giller, A. G. Hildrew & D. G. Raffaelli), pp. 649. Oxford: Blackwell Scientific Ltd.

Hoey, T. B., Smart, D. W. J., Pender, G. & Metcalfe, N. 1998 Engineering Methods fo r Scottish Gravel Bed Rivers. SNH Review No. 47. Battleby: Scottish Natural Heritage.

Holomuzki, J. R. 1996 Effects of substrate and predator type on microdistributions and drift of a lotic mayfly. Journal o f the North American Benthological Society 15, 520-528.

Hughes, J. M., Bunn, S. E., Hurwood, D. A. & Cleary, C. 1998 Dispersal and recruitment of Tasiagma ciliata (Trichoptera: Tasimiidae) in rainforest streams, south-eastern Australia. Freshwater Biology 39, 117-127.

Hughes, J. M., Mather, P. B., Sheldon, A. L. & Allendorf, F. W. 1999 Genetic structure of the stonefly, Yoraperla brevis, populations: the extent of gene flow among adjacent montane streams. Freshwater Biology 41. 63-72.

Huhta, A., Muotka, T. & Tikkanen, P. 1995 Diel foraging periodicity of lotic mayfly (Ephemeroptera) nymphs during the sub-arctic summer. Archiv Fur Hydrobiologie 134, 281-294.

Hurlbert, S. H. 1984 Pseudoreplication and the design of ecological field experiments. Ecological Monographs 54, 187-211.

Huryn, A. D. & Wallace, J. B. 1988 Community structure of Trichoptera in a mountain stream - spatial patterns of production and functional-organization. Freshwater Biology 20, 141-155.

Hutchinson, G. E. 1957 Geography, physics and chemistry. A Treatise on Limnology. New York: John Wiley & Sons, inc.

Hynes, H. B. N. 1972 The Ecology o f Running Waters. Liverpool: Liverpool University Press.

Hynes, H. B. N. 1976 Biology of Plecoptera. Annual Rview o f Entomology 21, 135-153.

195

References

Hynes, H. B. N., Williams, D. D. & Williams, N. E. 1976 Distribution of the benthos within the substratum of a Welsh mountain stream. Oikos 27, 307-310.

Jenkins, R. A., Wade, K. R. & Pugh, E. 1984 Macroinvertebrate habitat relationships in the Tefi catchment and significance to conservation. Freshwater Biology 14, 23-42.

Jeppesen, E., Iversen, T. M., Sand-Jensen, K. & Jorgensen, C. P. 1984 0kologiske konsekvenser af reduceret vandforing i Sus&n. In Biologiske processor og vandvalitets-forhold., vol. 2. Copenhagen: Miljostryrelsen.

Johnson, I. W., Elliott, C. R. N. & Gustard, A. 1995 Using the IFIM to model salmonid fish habitat in the River Allen, England. Bulletin Francais De La Peche Et De La Pisciculture , 355-363.

Johnson, I. W. & Law, F. M. 1995 Computer-models for quantifying the hydro-ecology of British rivers. Journal o f the Chartered Institution o f Water and Environmental Management 9, 290-297.

Jones, N. V., Litterick, M. R. & Pearson, R. G. 1977 Stream flow and the behaviour of caddis larvae. In Proceedings o f the Second International Conference on Trichoptera. Junk, The Hague.

Jongman, R. H. G., Ter Braak, C. J. F. & Van Tongeren, O. F. R. 1988 Data Analysis in Community and Landscape Ecology. Cambridge: Cambridge University Press.

Jowett, I. G. 1993 A method for objectively identifying pool, run, and riffle habitats from physical measurements. New Zealand Journal o f Marine and Freshwater Research 27, 241-248.

Jowett, I. G. 1998 Hydraulic geometry o f New Zealand rivers and its use as a preliminary method of habitat assessment. Regulated Rivers-Research & Management 14, 451-466.

Jowett, I. G. & Richardson, J. 1995 Habitat preferences of common riverine New Zealand native fishesand implications for flow managment. New Zealand Journal o f Marine and Freshwater Research 29, 13-23.

Jowett, I. G., Richardson, J., Biggs, B. J. F., Hickey, C. W. & Quinn, J. M. 1991 Microhabitat preferences of benthic invertebrates and the development of generalized Deleatidium spp habitat suitability curves, applied to 4 New-Zealand rivers. New Zealand Journal o f Marine and Freshwater Research 25, 187-199.

Jowett, I. G. 1997 Instream flow methods: A comparison of approaches. Regulated Rivers-Research & Management 13, 115-127.

Keckeis, H. & Schiemer, F. 1992 Food-consumption and growth of larvae and juveniles of 3 cyprinid species at different food levels. Environmental Biology o f Fishes 33, 33-45.

Kiel, E. & Frutiger, A. 1997 Behavioural responses of different blackfly species to short -term oxygen depletion. International Revue der gesampten Hydrobiologie 82, 107-120.

Koetsier, P. & Bryan, C. F. 1995 Effects of abiotic factors on macroinvertebrate drift in the lower Mississippi river, Louisiana. American Midland Naturalist 134, 63-74.

Kohler, S. L. 1985 Identification of stream drift mechanisms: An experimental and observational approach. Ecology 66, 1749-1761.

196

References

Kratz, K. W. 1996 Effects of stoneflies on local prey populations - mechanisms of impact across prey density. Ecology 77, 1573-1585.

Lacoursiere, J. O. 1991 A laboratory study of fluid flow and microhabitat selection by larvae of Simulium vittatum (Diptera: Simuliidae). Canadian Journal o f Zoology 70, 582-596.

Lamouroux, N. 1993 Comment on 'The nature of the fluid boundary layer and the selection of parameters for benthic ecology' (Carling, 1992). Freshwater Biology 30, 331-332.

Lancaster, J. 1992 Diel variations in the effect of spates on mayflies (Ephemeroptera, Baetis). Canadian Journal o f Zoology 70, 1696-1700.

Lancaster, J. 1999 Small-scale movements of lotic macroinvertebrates with variations in flow. Freshwater Biology 41, 605-619.

Lancaster, J. & Hildrew, A. G. 1993a Characterising in-stream flow refugia. Canadian Journal o f Fisheries and Aquatic Sciences 50, 1663-1675.

Lancaster, J. & Hildrew, A. G. 1993b Flow refugia and the microdistribution of lotic macroinvertebrates. Journal o f the North American Benthological Society 12, 385-393.

Lancaster, J., Hildrew, A. G. & Townsend, C. R. 1990 Stream-flow and predation effects on the spatial dynamics of benthic invertebrates. Hydrobiologia 203, 177-190.

Langan, S. J., Wade, A. J., Smart, R., Edwards, A. C., Soulsby, C., Billett, M. F., Jarvie, H. P., Cresser,M. S., Owen, R. & Ferrier, R. C. 1997 The prediction and management of water quality in a relatively unpolluted major Scottish catchment: Current issues and experimental approaches. Science o f the Total Environment 194, 419-435.

Layzer, J. B. & Madison, L. M. 1995 Microhabitat use by ffesh-water mussels and recommendations for determining their instream flow needs. Regulated Rivers-Research & Management 10, 329-345.

Leader, J. P. 1976 Marine caddis flies (Trichoptera: Philanisidae). In Marine Insects (ed. L. Cheng), pp. 581. Amsterdam: North-holland publishing company.

LeRoy Poff, N. 1996 A hydrogeography of unregulated streams in the United States and an examination of scale-dependence in some hydrological descriptors. Freshwater Biology 36, 71-91.

Levins, R. 1969 Some demographic and genetic consequences of environmental heterogeneity for biological control. Bulletin o f the Entomological Society o f America 15, 237-240.

Lynch, J. M. & Murray, D. A. 1994 Fishery rehabilitation enhancment following arterial drainage in Ireland. Verh. Internat. Verein. Limnol. 25, 1502-1508.

Macan, T. T. 1963 Freshwater Ecology: Longmans, Green & Co. Ltd.

Macan, T. T. 1977 A Key to the British Fresh- and Brackish-water Gastropods with Notes on theirEcology. Scientific Publication No. 13. Far Sawrey, Ambleside, Cumbria: Freshwater Biological Association.

Maitland, P. S. 1977 A Coded Checklist o f Animals Occurring in Fresh Water in the British Isles:Institute of Terrestrial Ecology.

197

References

Maitland, P. S. 1990 Biology o f Freshwater s. Tertiary Level Biology. Glasgow & London: Blackie.

Makipaa, R. 1999 Response patterns of Vaccinium myrtillus and N-vitis-idaea along nutrient gradients in boreal forest. Journal o f Vegetation Science 10, 17-26.

Malicky, H. 1981 The phenology o f dispersal of several caddisfly (Trichoptera) species in the island of Crete. In Third International Symposium on Trichoptera. (ed. G.P. Moretti). Junk The Hague, 157-163.

Malmquist, B. & Otto, C. 1987 The influence of substrate stability on the compositon of stream benthos: an experimental study. Oikos 48, 33-38.

Marshall, E. J. P. & Westlake, D. F. 1990 Water velocities around water plants in chalk streams. Folia geobotanic et phytotaxonomica 25, 279-289.

Matthaei, C. D., Uehlinger, U. & Frutiger, A. 1997 Response of benthic invertebrates to natural versus experimental disturbance in a Swiss prealpine river. Freshwater Biology 37, 61-77.

Matthaei, C. D., Uehlinger, U., Meyer, E. I. & Frutiger, A. 1996 Recolonization by benthic invertebrates after experimental disturbance in a swiss prealpine river. Freshwater Biology 35, 233-248.

McCracken, D.I. 1994 A fuzzy classification o f moorland ground beetle (Coleoptera: Carabidae) and plant communities. Pedobiologia 38 12-27.

McCune, B. 1997 Influence of noisy environmental data on canonical analysis. Ecology 78, 2617-2623.

McKirdy, A. 1999^4 Landscape Fashioned by Geology: Scotland - Creation o f its Natural landscape'. Scottish Natural Heritage.

McShaffrey, D. & McCafferty, W. P. 1987 The behaviour and form of Psephenus herrichi(DeKay) (Coleoptera: Psephenidae) in realtion to water flow. Freshwater Biology 18, 319-324.

Merrit, R. W. & Cummins, K. W. 1979 An Introduction to Aquatic Insects. Dubuque, Iowa: Kendall/Hunt Publishing.

Milhous, R. T., Wegner, D. L. & Waddle, T. 1984 User's guide to the physical habitat simulation system. US Fish and Wildlife Service Biological Services Program FWS/OBS-81/43.

Miller, P. 1987 Dragonflies. Naturalists' handbooks 7. Cambridge: Cambridge University Press.

Minshall, G. W. 1984 Aquatic insect -substratum relationships. In Ecology o f Aquatic Insects.

Minshall, G. W., Cummins, K. W., Petersen, R. C., Cushing, C. E., Bruns, D. A., Sedell, J. R. &Vannote, R. L. 1985 Developments in stream ecosystem theory. Canadian Journal o f Fisheries and Aquatic Sciences 42. 1045-1055.

Moog, D.B. & Jirka, G.H. 1999 Stream reaeration in nonuniform flow: Macroroughness enhancement. Journal of Hydraulic Engineering-ASCE, Vol. 125, N o.l , pp. 11-16

Morgan, N. C. & Egglishaw, H. J. 1966 A survey of the bottom fauna of streams in the Scottish Highlands. Pt. I. Composition of the fauna. Hydrobiologia 25, 181-211.

Moss, B. 1988 Ecology o f Freshwaters Man and Medium. Oxford: Blackwell Scientific Publications.

198

References

Nagell, B. & Larshammar, P. 1981 Critical oxygen-demand in Plecoptera and Ephemeroptera nymphs as determined by 2 methods. Oikos 36, 75-82.

Newbury, R. & Gaboury, M. 1993 Exploration and rehabilitation of hydraulic habitats in streams using principles of fluvial behaviour. Freshwater Biology 29, 195-210.

Nowell, A. R. M. & Church, M. 1979 Turbulent flow in a depth-limited boundary layer. Journal o f Geophysical Research 84, 4816-4823.

O'Connor, N. A. 1991 The effects of habitat complexity on the macroinvertebrates colonising wood substrates in a lowland stream. Oecologia , 504-512.

Ormerod, S. J. 1988 The micro-distribution of aquatic macroinvertebrates in the Wye river system: the result of biotic or abiotic factors? Freshwater Biology 20, 241-247.

Otto, C. 1976 Factors affecting the drift o f Potamophylax cingulatus (Trichoptera) larvae. Oikos 27, 93- 100.

Palmer, C., O Keeffe, J. & Palmer, A. 1993a Macroinvertebrate functional feeding groups in the middle and lower reaches of the Buffalo River, Eastern Cape, South-Africa .2. Functional-morphology and behaviour. Freshwater Biology 29, 455-462.

Palmer, C., O Keeffe, J., Palmer, A., Dunne, T. & Radloff, S. 1993b Macroinvertebrate functional feeding groups in the middle and lower reaches of the buffalo river, Eastern Cape, South-Africa .1. dietary variability. Freshwater Biology 29, 441-453.

Peeters, E. & Gardeniers, J. J. P. 1998 Logistic regression as a tool for defining habitat requirements of two common gammarids. Freshwater Biology 39, 605-615.

Petts, G., Maddock, I., Bickerton, M. & Ferguson, A. J. D. 1995 Linking Hydrology and Ecology: The scientific basis of river managment. In The Ecological Basis o f River Management (ed. D. M. Harper & A. J. D. Ferguson), pp. 1-16. Chichester: John Wiley & Sons.

Petts, G. E. & Bickerton, M. A. 1994 Influence of water abstraction on the macroinvertebrate community gradient within a glacial stream system: La Borgne d'Arolla, Valais, Switzerland. Freshwater Biology 32, 375-386.

Poff, N. L., Palmer, M. A., Angermeier, P. L., Vadas, R. L., Hakenkamp, C. C., Bely, A., Arensburger, P. & Martin, A. P. 1993 Size structure of the metazoan community in a piedmont stream.Oecologia 95, 202-209.

Poff, N. L. & Ward, J. V. 1991 Drift responses of benthic invertebrates to experimental streamflowvariation in a hydrologically stable stream. Canadian Journal o f Fisheries and Aquatic Sciences 48, 1926-1936.

Pringle, C. M., Naiman, R. J., Bretschko, G., Karr, J. R., Oswood, M. W., Webster, J. R., Welcomme, R. L. & Winterboum, M. J. 1988 Patch dynamics in lotic systems: the stream as a mosaic. Journal o f the North American Benthological Society 7, 503-524.

Puckridge, J. T., Sheldon, F., Walker, K. F. & Boulton, A. J. 1998 Flow variability and the ecology of large rivers. Marine and Freshwater Research 49, 55-72.

199

References

Quinn, J. M. & Hickey, C. W. 1994 Hydraulic parameters and benthic invertebrate distributions in 2 gravel-bed New-Zealand rivers. Freshwater Biology 32, 489-500.

Quinn, J. M., Hickey, C. W. & Linklater, W. 1996 Hydraulic influences on periphyton and benthicmacroinvertebrates - simulating the effects of upstream bed roughness. Freshwater Biology 35, 301-309.

Raddum, G. G. & Fjellheim, A. 1993 Life-cycle and production of Baetis rhodani in a regulated river in western Norway - comparison - o f pre-regulation and post-regulation conditions. Regulated Rivers-Research & Management 8, 49-61.

Reynolds, C. S. 1998 The state of freshwater ecology. Freshwater Biology 39, 741-753.

Rincon, P. A. & LobonCervia, J. 1997 Temporal patterns in macroinvertebrate drift in a northern Spanish stream. Marine and Freshwater Research 48, 455-464.

Rosillon, D. 1988 Food preference and relative influence of temperature and food quality on life-history characteristics of a grazing mayfly, Ephemerella-ignita (poda). Canadian Journal o f Zoology 66, 1474-1481.

Rossaro, B. & Pietrangelo, A. 1993 Macroinvertebrate distribution in streams: a comparision of CA ordination with biotic indices. Hydrobiologia 263, 109-118.

Rutt, G. P., Weatherley, N. S. & Ormerod, S. J. 1989 Microhabitat availibility in Welsh moorland and forest streams as a determinant of macroinvertebrate distribution. Freshwater Biology 22, 247- 261.

Sabbatini, M. R. & Murphy, K. J. 1996 Response of Callitriche and Potamogeton to cutting, dredging and shade in English drainage channels. Journal o f Aquatic Plant Management 34, 8-12.

Saetersdal, M. & Birks, H. J. B. 1997 A comparative ecological study of Norwegian mountain plants in relation to possible future climatic change. Journal o f Biogeography 24, 127-152.

Sand-Jensen, K. 1998 Influence of submerged macrophytes on sediment composition and near- bed flow in lowland streams. Freshwater Biology 39, 663-679.

Sand-Jensen, K. & Madsen, T. V. 1992 Patch dynamics of the stream macrophyte, Callitriche- cophocarpa. Freshwater Biology 27, 277-282.

Sand-Jensen, K. & Mebus, J. R. 1996 Fine-scale patterns of water velocity within macrophyte patches in streams. Oikos 76, 169-180.

Scarsbrook, M. R. & Townsend, C. R. 1993 Stream community structure in relation to spatial andtemporal variation: a habitat template study of two contrasting New Zealand streams. Freshwater Biology 29, 395-410.

Schiemer, F. & Wieser, W. 1992 Food and feeding, ecomorphology, energy assimilation and conversion in cyprinids - epilogue. Environmental Biology o f Fishes 33, 221-221.

Schmidt - Nielson, K. 1990 Animal Physiology. Adaptation and Environment. Cambridge: Cambridge University Press.

200

References

Sidney, S. 1956 Nonparametric Statistics for the Behavioural Sciences. Tokyo: McGraw-Hill Book Company, Inc.

Smith, I. R. 1975 Turbulence in Lakes and Rivers. Scientific publication no. 29. Ambleside: Freshwater Biological Association.

Smith, I. R. & Lyle, A. A. 1994 Running Waters. In The Freshwaters o f Scotland, A National Resource o f International Significance (ed. P. S. Maitland, P. J. Boon & D. S. Mclusky), pp. 17-34: John Wiley & Sons.

Smith, J. A. & Dartnall, A. J. 1980 Boundary layer control by water pennies (Coleoptera: Psephenidae).Aquatic Insects 2, 65-72.

Southwood, T. R. E. 1977 Habitat: the template for ecological strategies? Journal o f Animal Ecology 46, 337-365.

Southwood, T. R. E. 1988 Tactics, strategies and templates. Oikos 52, 3-18.

Statzner, B. 1988 Growth and Reynolds-number o f lotic macroinvertebrates - a problem for adaptation of shape to drag. Oikos 51, 84-87.

Statzner, B. & Higler, B. 1985 Questions and comments on the river continuum concept. Canadian Journal o f Fisheries and Aquatic Sciences 42, 1038-1044.

Statzner, B. & Higler, B. 1986 Stream hydraulics as a major determinant of benthic invertebrate zonation patterns. Freshwater Biology 16, 127-139.

Statzner, B. & Holm, T. F. 1982 Morphological adaptations of benthic invertebrates to stream-flow - an old question studied by means of a new technique (laser doppler anemometry). Oecologia 53, 290-292.

Statzner, B. & Holm, T. F. 1989 Morphological adaptation of shape to flow: Microcurrents around lotic macroinvertebrates with known Reynolds numbers at quasi-natural flow conditions. Oecologia 78, 145-157.

Strevens, A. P. 1999 Impacts of groundwater abstraction on the trout fishery of the River Piddle, Dorset; and an approach to their alleviation. Hydrological Processes 13, 487-496.

Surber, E. W. 1937 Rainbow trout and bottom fauna in one mile of stream. Transactions o f American Fisheries Society 66, 193-202.

Survey, G. 1971 Geological map of Great Britain. Sheet 1. Chessington. Ordance Survey.

Survey, G. 1977 Geological Survey, North Sheet: Institute of Geological Sciences.

Svensson, B. W. 1972 Flight periods, ovarian maturation and mating in Trichoptera at a South Swedish stream. Oikos 23, 370-383.

Swales, S. & O'Hara, K. 1980 Instream habitat improvement devices and their use in freshwater fisheries management. Journal o f Environmental Management 10, 167-179.

ter Braak, C. J. F. & Prentice, I. C. 1988 A theory of gradient analysis. Advances in Ecological Research 18. 271-317.

201

References

ter Braak 1997, 1998 Canoco for Windows. Wageningen.

ter Braak, C. J. F. & Smilauer, P, 1998 CANOCO Reference Manual and User's Guide to Canoco for Windows. Ithaca: Microcomputer Power.

Thorup, J. 1966 Substrate type and its value as a basis for the delimitation of bottom fauna communities in running waters. Special publication number4 Pymatuning Labortory o f Ecology University o f Pittsburg , 59-74.

Tikkanen, P., Muotka, T. & Huhta, A. 1994 Predator detection and avoidance by lotic mayfly nymphs of different size. Oecologia 99, 252-259.

Tockner, K. & Schiemer, F. 1997 Ecological aspects of the restoration strategy for a river-floodplainsystem on the Danube River in Austria. Global Ecology and Biogeography Letters 6, 321-329.

Tokeshi, M. 1993 On the evolution of commensalism in the Chironomidae. Freshwater Biology 29, 481- 489.

Tokeshi, M. 1994 Community ecology and patchy freshwater habitats. In Aquatic Ecology: Scale, Pattern and Process (ed. P. S. Giller, A. G. Hildrew & D. G. Raffaelli), pp. 63-92. Oxford: Blackwell Science.

Townsend, C. R. 1980 The Ecology o f Streams and Rivers. Studies in Biology no. 122. Southhampton: Camelot Press Ltd.

Townsend, C. R. 1989 The patch dynamics concept of stream community ecology. Journal o f the North American Benthological Society 8, 36-50.

van't Woudt, B. D. & Nicolle, K. 1978 Flow processes below a gravelly riverbed. Journal o f Hydrology (N.Z.) 17, 103-120.

Vogel, S. 1994 Life in Moving Fluids The Physical Biology o f Flow. Princeton, New Jersey: Princeton University Press.

Walker, P. M. B. 1995 Dictionary o f Science and Technology. Edinburgh & New York: Larousse.

Wallace, I. D. 1981 A key to larvae of the family Leptoceridae (Trichoptera) in Great- Britain and Ireland.Freshwater Biology 11, 273-297.

Wallace, I. D., Wallace, B. & Philipson, G. N. 1990 4 Key to the Case Bearing Caddis Larvae o f Britain and Ireland. Scientific Publication No. 53. Far Sawrey: Freshwater Biological Association.

Walton Jr, O. E. 1978 Substrate attachment by drifting aquatic insect larvae. Ecology 59, 1023-1030.

Ward, J. V., Bretschko, G., Brunke, M., Danielopol, D., Gibert, J., Gonser, T. & Hildrew, A. G. 1998 The boundaries of river systems: the metazoan perspective. Freshwater Biology 40, 531-569.

Waringer, J. A. 1987 Spatial-distribution of Trichoptera larvae in the sediments of an Austrian mountain brook. Freshwater Biology 18, 469-482.

Waringer, J. A. 1989a Life cycle, horizontel microdistribution and current resistance of Allogamusauricollis (Trichoptera: Limnophilidae) in an Austrian mountain brook. Freshwater Biology 22, 177-188.

202

References

Waringer, J. A. 1989b Resistance of a cased caddis larva to accidental entry into the drift - the contribution of active and passive elements. Freshwater Biology 21, 411-420.

Waringer, J. A. 1993 The drag coefficient of cased caddis larvae from running waters - experimental- determination and ecological applications. Freshwater Biology 29, 419-427.

Webb, B. W. & Walling, D. E. 1993 Temporal variability in the impact of river regulation on thermal regime and some biological implications. Freshwater Biology 29, 167-182.

Weissenberger, J., Spatz, H., Emanns, A. & Schwoerbel, J. 1991 Measurement of lift and drag forces in the mN range experienced by benthic arthropods at flow velocities below 1.2 m/s. Freshwater Biology 25, 21-31.

Welcomme, R. L. 1992 River conservation - Future perspectives. In River Conservation and Management (ed. P. J. Boon, P. Calow & G. E. Petts), pp. 453-462. Chichester: John Wiley & Sons.

Wells, A. 1992 The 1st parasitic Trichoptera. Ecological Entomology 17, 299-302.

Werritty, A., Brazier, V., Gordon, J. E. & McManus, J. 1994 Geomorphology. In The Freshwaters o fScotland, A National Resource o f International Significance (ed. P. S. Maitland, P. J. Boon & D. S. Mclusky), pp. 65-88: John Wiley & Sons.

Whetmore, S. H., Mackay, R. J. & Newbury, R. W. 1990 Characterization of the hydraulic habitat of Brachycentrus - occidentalis, a filter-feeding caddisfly. Journal o f the North American Benthological Society 9, 157-169.

Williams, D. D. & Feltmate, B. W. 1994 Aquatic Insects. Wallingford: CAB International.

Williams, D. D. & Smith, M. R. 1996 Colonisation dynamics of river benthos in response to local changes in bed charcterstics. Freshwater Biology 36, 237-248.

Williams, D. D., Tavares, A. F. & Bryant, E. 1987 Respiratory device or camouflage - a case for the caddisfly. Oikos 50, 42-52.

Wohl, D. L., Wallace, J. B. & Meyer, J. L. 1995 Benthic macroinvertebrate community structure, function and production with respect to habitat type, reach and drainage-basin in the southern Appalachians (USA). Freshwater Biology 34, 447-464.

Wohl, E. E., Vincent, K. R. & Merritts, D. J. 1993 Pool and Riffle Characteristics in Relation to Channel Gradient. Geomorphology 6, 99-110.

Wright, J. F. 1992 Spatial and temporal occurrence of invertebrates in a chalk stream,. Hydrobiologia 248, 11-30.

Wright, J. F., Blackburn, J. H., Westlake, D. F., Furse, M. T. & Armitage, P. D. 1995 Anticipating theconsequences of river management for the conservation of macroinvertebrates. In The Ecological Basis o f River Management (ed. D. M. Harper & A. J. D. Ferguson), pp. 137-150. Chichester: John Wiley & Sons.

Young, W. J. 1992 Clarification of the criteria used to identify near-bed flow regimes. Freshwater Biology 28, 383-391.

203

References

Young, W. J. 1993 Field techniques for the classification of near-bed flow regimes. Freshwater Biology 29, 377-383.

204

Appendix I

Appendix I: Definitions of flow units and parameters

Symbol Definition

D Water depth (m)Fr Froude number

g Acceleration due to gravity 9.8m2s'1

ks Nikuradse’s roughnessRe Reynolds numberRe* Reynolds roughness numberU Water velocity measured at 0.4D from the substrate. It equates to depth

averaged velocity in a full developed boundary layer.u* Shear velocity (ms'1)

Zo Characteristic Roughness length (m)

6 Thickness of the laminar sublayer (mm)

V kinematic viscosity (= 1 X 10'6 m V 1 at 20 °C)

To Boundary shear stress (Nm'2)

205

Appendix II

Appendix II: Log plotting of velocity profiles

Introduction

Plots of velocity against natural log of depth, show a linear relationship from which a

number of hydraulic variables can be estimated. Variables generated from this

relationship and presented in this thesis were produced using a Microsoft Excel

template file designed by Dr. Trevor Hoey of Glasgow University Department of

Geography and Topographic Sciences. A large number o f these files were generated

during analysis, River Etive (49), Duneaton Water (52) and Blane Water (42), see

appendix III. An example from one river is given showing the generation of derived

variables from raw data. For all other regressions summaries of the variables generated

are presented. Velocity profiles were chosen to be used in further analysis if their

regression lines were statistically significant. I checked for patterns in the data that

may bias further analysis.

Methods

When the velocity profile is plotted on a lognormal co-ordinate shear velocity (u*) and

equivalent bed hydraulic roughness (k s ) can be calculated.

Shear velocity is inversely proportional to the gradient o f the velocity profile and

z 0 (characteristic roughness length) can be calculated by extrapolating the regression

line to zero velocity. The variable z 0 can be taken as an estimate of k s . These

variables are useful in classifying near bed flow. Reynolds roughness number (Re*)

can be calculated from these two variables.

Re* = u+ks / v

206

Appendix II

Critical values of Re* delineate flow as to the presence of the laminar sublayer of the

boundary layer. Where the laminar sublayer is present its thickness ( J ) can be

estimated as follows:

S = 11.6v / m*

Working Example: Sample point Blane Water Section A, Transect 1, Sample A.

(BLAT1SA)

Height above bed (m)

In (ht. Above bed)

Velocity(m/s)

0.01 -4.61 0.260.012 -4.42 0.280.014 -4.27 0.300.018 -4.02 0.210.02 -3.91 0.370.04 -3.22 0.410.06 -2.81 0.460.08 -2.53 0.490.1 -2.30 0.520.12 -2.12 0.51

The plot of log depth against velocity is visually checked for a linear relationship

between the two variables, Figure ii-1. In this case the left most point is an obvious

outlier and does not follow the line described by the other points but it is not removed

from the analysis. The linear relationship between In depth and velocity in fully

developed boundary layers is a well described in the literature and is an accepted fact

by both physicists, hydrologists and geomorphologists hence it is acceptable to remove

rogue points at this stage that deviate from a clear pattern. It has been suggested that

the estimation of bed shear stress from velocity profiles in a shallow river environment

can be improved by using only the section of the profile from close to the bed (Biron

etal. 1998).

207

Appendix II

In a number of cases (R.Etive 14, Duneaton Water 2, Blane Water 0) I did remove

points which did not follow the log-linear relationship. This was exclusively where

points high up in a profile failed to follow the log-linear relationship, Figure ii-2. In the

example plotted in Figure ii-2 the removal of the top 4 points of the plot improved the

fit from R2 = 0.31 to R2 = 0.87.

There are a number of situations were velocity profiles of this type are likely to

happen, ' ...non-logarithmic layers may occur in strongly accelerating or de-

accelerating flow such as occurs in river reaches which narrow or widen rapidly. The

logarithmic profile may be distorted by spatial variation in the mixture o f roughness

types on the bed, or by extreme bed roughness such as occurs over dunes and over

large rocks. ..... However even in these circumstances it is possible to obtain near

log-normal distributions o f velocity, from close to the bed, which can be used to

estimate the local bed shear stress ' from (Carling 1992). The second instance where

bed roughness is the important factor would occur at the 3 rivers examined in this

study.

0.00 -|

T3<]> -1.00 -<u& -2.00 -cd

43OX) -3.00 -(D

43e -4.00 -

-5.00 -

0.00 0.10 0.20 0.30 0.40 0.50 0.60

Velocity (m /s)

Figure ii-1 Velocity (u) measured at a number of depths above a single point on the stream bed plotted against In depth (D).

208

Appendix II

•§> -3.00

0.00 0.05 0.10 0.15 0.20

Velocity (m /s)

0.25 0.30

Figure ii-2 Plot from River Etive data, section A, Transect 4, sample c showing the breakdown of log linear relationship between depth and velocity higher up the water column.

Results

Estimates of the variables generated for our working example are given in table ii-2:

Table ii-2, Estimates of flow parameters from log normal plot of velocity against depth for River Etive sample A4C.

Variable Estimate

rA2 0.87Gradient 0.11Intercept 0.77

zo 0.001136u* 0.0454

tau 2.06

s.e. grad 1.67E-02

s.e. tau 6.07E-01Umax 0.53

U profile 0.42U real 0.42U at 0.37 depth 0.42Fr 0.38

The R-squared value for each line was checked for its significance by looking up

standard statistical tables. There was a clear relationship between the significance of

profiles and mean water column velocity. The ranked plots showing the distribution of

209

Appendix II

significant velocity profiles to mean water column velocity show the majority of non

significant profiles occurred in slow water in the Blane water and the Duneaton water

Figures ii-3-5. This pattern was also evident in the River Etive but here there was also

a strong scatter o f non-significant profiles over the entire range of velocities sampled.

Blane

1.6 T

1.4 -

+ P 0.01 QN/s0.8

♦ ♦ ♦0.6 -

0.4 -

0.2

20 25 30 35

Samples

Figure ii-3 Rank order of Blane Water samples (velocity profiles) by increasing velocity mean water column velocity. Diamonds = statistically significant velocity profiles, squares = non significant velocity profiles.

Duneaton

0.7 T

0.6e>i 0.5

o 0.4om> 0.3cra® 0.2 -

0 10 20 30 40 50 60

Samples

Figure ii-4 Rank order of Duneaton Water samples (velocity profiles) by increasing velocity mean water column velocity. Diamonds = statistically significant velocity profiles, squares = non significant velocity profiles.

210

Appendix II

E tive

0.8 T

0.7 -

2* 0.5 u•2 0.4

30

Samples

Figure ii-5 Rank order of River Etive samples (velocity profiles) by increasing velocity mean water column velocity. Diamonds = statistically significant velocity profiles, squares = non significant velocity profiles.

Discussion

The method worked well. The observed bias o f non significant velocity profiles

occurring in low velocity areas agrees with field observations. The most difficult

velocity profiles to measure were in the slow shallows but also in the deeper water

when slow. This often occurred behind large substrate elements where one would

expect low flows. In the R. Etive large boulders frequently acted as weirs causing

areas o f both decelerating and accelerating water. The data presented here in

combination with substrate measures can be used to identify the qualitative flow

categories of skimming, wake interference and isolated roughness flows, but this

aspect was not carried out (Davis & Barmuta 1989; Nowell & Church 1979).

211

<n'inroc(0coin<no>k.U)0)k.

■Oc(0<Q*-«C0

T3S(0k.

<n0)

o*_a>

o0 a>>

== c— ox *;1 =I ?a.< -E

cc3hj03

T3

03(_CD

GH

jdgo3

G£o

03G45

Go03

w

tHou,Oh03<L>

G03

J 53

<dUh45

G£

C'oCLh00G

3 ,So3(Z)

3

<2CAJJ23ot_O h

oo45>45

G

co3H->ooo

33a<L>O hO hc3ca

3H

VP h

occ3o

t oO.00

"55+-»OC

CA

3

bG-t-j'5<dV-Hcc3caa>

3>Ph

H—>ao

3>

C"3oo03

45G45>o

Go3+-•00

"5

O h45

T300O

t+Ho

3g03co

45Uh0045Uht-H

3175o

03H-<CA

45GH-JG-*->' ?T3(Oh->C<0ca(OUhO h

GOo'IIVP hCOknCO

G£

T3s

OOOOVP h

ooo'VP h

oo

«Ak.45kHk*

0 s■ t

CA• pkG

45 kk

O 45kkH-* CD ’vi

■MCAkt

H-Su

Vi

Ge45

45CAG

kHkk66

kn 66« t A »kn jjj

5 s■QC CO WD o *£V 4)CA

.9 ««ID k.45 £k. Sce a>

4)G

cV

G66

5oua

45J3

>45>66

JCk*OG

“g I sw k - - O. ca

• P4u_ 450 G

13 ~> -a c

-o C6 4/ H-»G u4/a/ vs G c k- e6

> * 5 “G -S 66 **W5 2 H-» ^G ««.op âj

k.£ *-m aj66 £

1 ij= ®a >e4/ +*

Q

£ oa _,

73k- g v o

CJ^ «505 cnc *-« b pa c

- «

.a S

i =

« G G «« k. V G G 45 ^t : b45 *"

&

x 45CD 2

. ? 66

45

G 45

I IG Hk 00 O

' GCACA

- S 4fH **■G

i uHH *2M "5hH k -45 J 'G -G66 “H II

G O G

£ .^ 145 45* 3CA 4545 >ts b o 7 : J- Ga

£45 45O G45 •£

k. G 45 45 Ot :

£ ^ o 13 tS 3G (A O G

03"&APh

(OD

1 3>PP

crCA

P h_CAG 'O’ O 03CAWho3<o

P h

PhO

CO

1G

JO'O h6oJ

CD

o<N

o*Y~)

or- oto

oo(N

o ’ o ’ o o' o

OO o<Ncr>

o*rs o00CO

o 00 r—HO o ’ O o ’ o ’

O00O

oOS■rO<Nr-H

o00to

oto<N

oGr-o

O o ’ o ’ o ’ o ’ o o'

OVOo

oNO oOsO

o<Nto

o<NCN

or-

oo

o' o o' o' o ’ o o'

oo

o oNOo

oro

o o oGO

o' o o' o ’ o' o o'

o<No

or-m

o<No

00ooo

o<NT—4oGG

o(No

o' O o' o’ o' o' o’

00o

<N<N

00o ro

or"o

ooooo

o' o’ o' o’ o’ o o’

■ro

OsOS<N

ro

CO

oOO

O00G

ro

o o' o' o' o' o' o’

<NO

Ooo<N(Nr* 4o

NOCOo o

(N <NO

o' o o' o' o' o' o'

oo

t-hG<N

oO

^ro

<No

GO o

oo' o’ o' o’ o’ o' o’

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

thV

eloc

ityD

epth

o o oo 00 o o(—s G G

G <N Tj-H/ <N (N

* * * ** * * ** * * *

r-~ 00 G00 Os Os Oso' o' o’ o'

ON

oJ

rttG

03G

00

o

o3G

03CA

(Nc3

G

Velo

city

0.134

0.1

83

0.235

0.2

49

0.284

0.4

00

0.450

0.5

70

0.590

0.

530

oONr -H

oO nd -

o04m

o00vo

oO n04

o04O '

oro

omCO

oO 'o

oVOc o

oONo

ON00O

oc n0 4

ooCO

ovo

oood -

om

oVOCO

or -H

0 4o o o ’ o ’ o ’ d o ’ d o ’ o ’ o ’ d d o ’ o ’ o ’ d d o ’

ooo

o04in

OOc o

ooO '

oOO0 4

VOVOO '

oo

r -H

ON04

ovoo

COCO04

oooo

OOO nO

oOO

o0 4d -

O0 4

oVOd*

o04

ovod-

o00

o ' o ’ d o ’ o ’ o ’ o ’ o ’ o ’ d o ’ d o ’ d d o ’ d o ’ o ’

od -r -H

oo-d -

om04

oO nVO

om04

ovoO '

omo

ONd"04

oino

04d-

ovoo

ON0 0o

o0 41—H

oVOCO

oooo

o0 4d "

oor -H

ovoCO

od -1—H

o ' o ’ o ’ d o ’ o ’ d d o ’ d o ’ d d d d d o ’ d d

OoT—H

oONc o

OO 'f—H

O

VO

oVO^ H

ovoVO

o0 4o

d -oo o

d -o

ONj—i od -o

0 4ooo

ovoo

00O '0 4

od -o

O nO n0 4

oOOo

od -CO

oo

o o ’ d o ’ d d d o ’ o ’ d o ’ d d o ’ o ’ d d o ’ d

oVOO

O04co

oo

oO 'd -

oO nO

oO 'm

m

ovoO '

0404o

mo

o0 4o

voO 'o

o0 4o

c nONT—H

o0 4o

voc n0 4

od -o

ON0 004

ovoo

o ’ O o ’ d O o ’ o ’ d o ’ d d o ’ o ’ d d d o ’ d d

o0 4o

r H

c o(N

o04O

0404ON

O0 4O

oCO

d -

omO '1—H

o0 4o

04m

0 0

od -O 'o

00

o00O n

0 0

of—H

0 4

o0 4O

O 'oo04

o0 4O

o o ’ o ’ o ’ o ' o ’ d o ’ o ’ o ' o ’ d o o ’ d o ’ o ’ o ' o ’

OOT—<o

ini n04

00

oc oO '04

0 0

O

omCO

CO

o00CO^ H

00

oom04

voo

mVOO

VO

oO nvo

voo

O '

0 4

00

o000004

00

oo ' o ’ d o ’ o ' o ’ d o ’ o ’ d o ’ o ’ o ’ d o ’ d o ’ o ’ o ’

d -1—H

oo-04

d -T“ H

oc nm04

d -^ -H

o0 0ON0 4

0 4

oooCO

d -

oONCO

"sf

o0 0O 'o

d-o

0 0VO

d -r -H

oON0 40 4

d "

or - '04

d"r -H

Oo ' o ’ d o ’ o ’ o ’ d o ’ o ’ d o ’ d o ’ o ’ o ’ o ’ o ’ d d

04

ocoON

04

oi nON04

0 4

oO noCO o

I— <

CO04

oON

0 4

or —1*

VOo

04

o04CO1 -H

0 4

od " 00 r H

04

o04O -0 4

04

Oo ’ d o ’ d o ' d o ’ o ’ o ’ o ‘ d o ’ o ’ d o ’ d d o ’ d

or H

o00o

oo

00d -» 4

oo

0 4OO

oo

ood-

or -H

oCOo

oo

ovoo

or—1o

ooo

04O nr —H

ot—H

oo0004

oo

o ’ d d d d o ’ d d d d d d d o ’ d d d o ’ d

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

Vel

ocity

Dep

th

oVi-t-*ed -*—>ViH-I

P h

crC/3

P4J /3

"S Ô edViWhCd

inON

edVimcd

-O

OOON

■d-

ooON

ooTTvoen

O'ON

ooooo04

VOON

VOVOrvj

VO

o'

d -vooid -

0400

oooONo0 4

VOOn

OO0 4VOm

o«On

ONinoo

VOVO

J=>Vic n~cdIS

oVicoed

J=>

edVid"edIS

-OVid --4->cd

3

cdViined

3

-OC/3intd3

ot/3ined3

cdViVOcd

X>

OOO(N0 4

VOON

X)ViVO•4—>cd3

Velo

city

0.132

0.1

44

0.162

0.1

65

0.149

0.2

86

0.350

0.4

00

0.430

0.

380

blat

6sc

9 0.9

6 **

* 18

2.00

Dept

h 0.0

10

0.012

0.0

14

0.018

0.0

20

0.050

0.1

00

0.150

0.

200

Sam

ple

Num

ber

of Pe

arso

n's

Rsq

P va

lue

F st

atist

ic

poin

ts ad

j

Velo

city

0.095

0.1

04

0.130

0.1

37

0.17

1 0.1

95

0.246

0.2

68

0.26

1 bl

at7s

a 9

0.89

***

62.57

De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.030

0.0

40

0.05

0

o o o o o X o o o X o o o o on - T—1 1—1 CO X X CO 0 0 CO F " CO CN 00 ■sj- Obr - H CO CN "St r -H CN CN CO CO CN o CO r -H X r H

o ’ o ' o ' o o ’ o ’ o ' o o ’ o ’ o ’ o ’ o o ’ o

CN o o o o o oo o o o CO oo O o o oX 1— I ^ H X CN CN CO r - X X X CN "St r - CO X’—1 I— I ^ t 1 CN r H CO CN CN o CO ^ H X I— I

O o o o o o ’ o ' o ' o ' o o o ’ o ’ o ’ o ’ o ’

o o CO o o o X O o o o X p ^ o o oX oo r~- CN Ob 00 I—I CN CO c - r - CN 00 CN CO oI— I o CN 1—1 CO o CN i— i CO ■—i ■—i o CN i—| X I— I

o ’ o ’ o ’ o ’ o ’ o ' o ’ O o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’

r - o o 's t o Ob o o o ■'sf 'S t Ob o o o■st x o r - o X O r - C" Ob CO CN oo X -St Xy—* o cn o CO o CN o CN o I— I O CN o X oo ' o ' o ' o ' o ’ o ’ o ’ o ’ o ' o ' o ’ o ’ o ’ o ’ o ’ o ’

c o o X o r - o 00 o X o CO o X o o oc o <N 00 CN X CN o CN ■St CN r - H CN 1— 1 CN X CNr - H o I—I o (N o CN o y—> o i—' o CO o X oo ' o o ’ o o ’ o ’ o ' o o ' o ' o ’ o ’ o o o ’ o ’

00 Ob 00 X 00 00 00 oo o 00 o 00 o oo1—1 i— i r - ^ h X r* H 1-H 1—1 Ob 1- o r - H ■St r -H CN 1—1I— I1 o ^ h o I— I1 o CN o I— I o ^ H o CN o X oo ’ o ' o ' o ’ o ’ o ' o ' o ' o ’ o ’ o o ’ o ’ o ’ o ’ o ’

o X -St X X X CN X r - X X X X o XI—1 r - H r - f—H CN I—1 00 1—1 X I— I CO i—i o 1— 1 r - 1— 1

i— ' o y—> o (N o r - H o y—> o i—| o CN o 'S t oo ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’

x CO 'vt CN ^ r X ■st Ob ^ r Ob 'S t t-" ■St o ■StO r “H X 1—H O i—i 00 I—1 X i—i CO i—i 'St I—I 1— 1 I—It “H o i— ' o CN o i—' o I—I1 o I—I o CN o 'S t oo ’ o ' o ' o ' O o ' o ' o o ’ o o o ’ o ’ o ’ o ’ o ’

X CN X CN CN CN CO CN o CN X CN oo CN o CNOb 1— 1 ^ r r— i—t i— i X r - H CN i—i 00 i-^ CO 1— 1 oo 1 - ^

o o •—1■ o CN o I—I o o o o CN o CO oo ’ o ' o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o

■St o X o ■st o o 00 o o r-- o o oo 1— 1 X 1—1 o l—l X r - H 1—1 1—1 o I— I CN 1—1 Ob 1—1I— I o ■—11 o CN o I— I o •—I o ^ H o CN o CO oo ’ o ’ o ’ o ' o ’ o ' o ' o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’

.ts 43 .ts -C .t i 43 .tJ ^3 .t i ^5 .t i -C .t iOJD(L>>

^ O<DQ £

o. g<Da go. o<D J ZQ ^ >

O- oo JxQ g >

a. g<DQ >

O h<DQ

CNCOx>Ob

Oboor-Ob

o oObCOoCO

oObr -CN

ox

o-stbd

oo

Obo'

Obo '

t-"f"- r-Ob

tor- obD COVO Ob

xtr>r-tg15

ocot"-+->egX

edco00tg3

XcoOO-*->eg

3

oCO00eg

3xx

x3

Velo

city

0.350

0.4

30

0.510

0.4

90

0.440

0.4

90

0.700

0.7

90

0.910

0.

950

blbt

2sa

10 -0

.11

n/s

0.08

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

40

0.060

0.0

80

0.10

0Ve

locit

y 0.0

27

0.027

0.0

36

0.087

0.0

68

0.132

0.0

68

0.115

0.

061

0.01

8

CNc oo

inc o

oCN

o oVOO

ovo

oc n

oo

oVOr — I

oo

^ rmo

VOmo

omo

CM

1—H

o" s f

oo00

oi no

or-~c n

oVOo

oCM

oVO

o ’ o o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o

oCOO

oCN

oo

oo

omo

oi nm

oor c

oo

ocnT“H

ovoo

CMc nO

CM o-3-o

r -o

oCM

ovooo

oo

oO

omo

oCN• r

o

o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ' o ’ o ' o ’ o ’ o ’ o ’ o ' o ’

0 0CNo

O s0 0

o0 0o

oo

o■'vl'o

o ot " -o

oo s

oo

o,!d -o

0 0CMo

oom

oc no

voi no

oO sO

o00r"-

mc nO

o0 0c n

o"xf-o

or -H

c n

oo

o ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ' o ’ o ’ o ’

CNo

T— 4 00

or ro

or foo

oc no

or -^ r

o^ ro

oc noo

ovoo

oc no

VOCMO

CM00

mCNo

O sVOO

oVOO

ocnt " -

Oc nO

ovoc n

oc no

O sCN

ovoo

o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’

o<No

oCNCN

oCNo

oVO00

oCNo

or -c n

oCMO

ooVO

oCNo

oCNOO

oCNo

c noor -H

oCNo

"d-o

OCMO

oVOm

OCNo

voCNc n

oCNo

r fO sCN

oCNo

C5 o ’ o ’ o o ’ o ' o ’ o o ’ o c> O o ' o o o ' o ’ o ’ o o ’ o ’

00 T H

oCOO s

00o

oo

00o

ooc n

00o

o00m

00o

o00

00o

voCM

ooo

omo

00o

o00

00o

oOc n

00o

T—H

O sCN

00o

o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ O o ’

VOI 1 4O

OS O

voo

ovo

vor —*o

CMooCN

voo

oId"VO

voo

o00r-

voo

OOT-Hvoo

c nc no

VO

oooo’d -

voo

C-O sCN

voo

oc n

voo

o ’ O o ’ o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’

ro

m 'If»—Ho

oi nC- o

c nvoCM o

o00m

ro

ovor - o

VO- ro

c no

ooO si n

ro

orc n

r—<o

O sooCM

^ r

oo ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’ o ' o ' o o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’

CN

oO sr**H

CN

ooOOvo

CN

ooc n

CMf 1O

omin

CM

oooVO

CNr -H

oO s^ r

CM

oc n■d-o

CN

O

oc nCM

CM

ovoc -CM

CM

oc n00CM

CM

Oo ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ' o ’ o ’ o ' o ’ o ’ o ’ o ’ o '

oo

m oo

ovo

oT“Ho

CNmCN

oo

oi n

oo

oc n

or -H

omo

oo

c -VOo

O

O

o oo

r -i nCN

o1—HO

O sVO

oo

o ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

th

ot/3o3on

Hh

CTC/3

P^J /3

"g -o*O «JC/5

c5<DPm

Cmov-i <«£x>

ooTTo

CMO s

O^ rvo’m

VOOO

oo00VOCM

r -ON

o■ rCMCM

Or "

ovoCM

C/5

'tS

ITSo '

oc n

r -

oC"rt-c o

O s

oCMOCN

ooVO

oC/3

CN-t-i£3

cciC/3

CO-t—' £ 3

X>C/3

CO■«->-O3

oC/3

CO£IB

ctfC/3

jO3

-Oc/3-+-*

3

oC/3

^ r-»—>jg3

ctiC/3

■*->jO3

C/3vn+->3

vo

VO

r'-~oo

oc/3m■<->3

Sam

ple

Num

ber

of Pe

arso

n's

Rsq

P va

lue

F st

atist

ic

poin

ts ad

jVe

locit

y 0.1

89

0.202

0.2

54

0.233

0.2

39

0.25

1 0.4

10

0.400

0.4

10

0.37

0bl

bt6s

a 10

0.82

** 42

.70

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

30

0.050

0.0

70

0.09

0Ve

locit

y 0.2

69

0.360

0.2

64

0.278

0.

251

0.340

0.4

60

0.460

0.5

30

0.58

0 o o o o o c n o o o oX oo O n CN X ON o c n CNo X O oo o CN t- h X o O n

o o ' o ' o ' o o ' o ’ o o ' o ’

o o O o o CN o o oo oX O n r - ON ON oo CN Xo X o 00 o CN o X o 00o o ' o ' o ’ o o o ’ o ' o ' o

o O o o o X o o X o<N X CN c n CN X c n CN O n

o X o I"- o CN o X o r ^ -

o ' o ’ o ’ o ' o ’ o ' o ’ o ’ o o

o o X o X T“ H o o oc n 00 c n X CN o r - CNo i n o X o CN o ^ r o Xo ' o ' o ' o ' o ’ o ’ o ’ o o ' o '

o o o o o " c r o o o oCN X CN X CN c n CN 00 CNO i n o r r o CN o CN o C "o o o ' o ’ o o ’ o ’ o ' o o ’

00 o 00 o 00 CN 00 o 00 or -H c n r —H X r -H oo r -H I— I Xo i n o X o CN o c n o C "o ' o ' o ' o ’ o o ' o ' o ’ o ’ o '

X o X o X c n X o X oi— i c n I— 1 c n I— I 1— 1 ^ r I— I r -H

o i n o X o CN o CN o r -o ' o ' o ' o ’ o o ’ o ’ o ' o ’ o

r3- o ^ r o TT I"- c n oI— 1 CN 1— 1 o I— I c n X r -H ooo i n o X o CN o CN o Xo o ' o ' o ’ o ' o ' o ' o ' o ' o ’

CN o CN o CN o CN CN CN oI—H 00 I— I CN I— I CN I— I X 1— 1 r— 1

o X o 'v f o CN o r -H o Xo ' o ' o ’ o o ’ o ' o ' o ' o ’ o '

o o o o o r - o CN o oI— I 1— 1 X I— I ON 1— 1 ’'nT I— I O no o o o CN o 1— 1 o T fo ' o ' o o ’ o ' o ’ o ' o ' o ’ o

js .t i x .t i j : .t ; . s a : .t iex<Da

oo1 3>

cx<DQ

oo13>

c x<L>Q

oo1 3>

cx(L)O

oo1 3>

c x<uQ

oo<d>

oXKc n

oooOnCN

CNc no '

o

ooo

ONC "

O nCN

cn'a

o00o ’

NOr -

00o CN

ONNOr-

x>cnNO+->jO3

ocnNO-t-*X3

aScnr-x3

xcnr"x3

or-+-»x3

216

o£ 4-S J S4-* u 4>4

+->VI©cn ?

U 3G ed 3

u 054J 4-1 1CfiS o> 3cn JC 4^

4^ cn4J"3 (3DO 3 tEej O4> 3 u

c .3 (31) -C ©A 4û 4) a9 i j *> 3«

■ o3(3D

4Ô33 s® —© © s«-> ^ -a

£ » -on f c ,45© £ IT) C45 £

a

* 1

Cd I —W B«W5 fc« I ^^ *S? ® *-£ *•» « ej

* s*5 " © ■©© 5^ CB■M_: u "O «45 cn

•© 3a j CdJs i

> *© nsa

u _ «£ •SP s

0 s* J8c T?s .is Q -cCN (U

1 --

45 £

a ©£ "3a jccn cn © 45« -O 2 *- «® s45 •"t£ 45

> p ~4- -®js *DD oW I a 73W oa cC•— cn 3 «.2° ^

osai u45£cnJDCOL.

J£ U3 ©4-»ed 4-»H ii

u3 > i.45cd£© u

0 -5£> * *5 "o© £ v u > 3

© «4 °£>=* £ edIS ©el- «« —* 3© ©+- .2 2 -fi© 4->

C©* ©NT -M O •■cnS . 2^ - 3c *■* © t!lV AA 2CUD 45

W €

O NO o m o CN o d - o d - o CN o m o00 O 00 CN CN r - NO On CN CN ON 00 in NOo o CN O CN o c n i—i o o o c n r - H CNo ' o ' o ' o ' o ' o ' o ’ O o ' o ' o ’ o ' o ’ o ’ o '

o o o NO o ON O 00 o NO o ON o r-~ oNO CN t - oo NO 00 o ON CN 00 o o d " oo o CN o I—1 o CN CN o o o o c n 1—1 CNo ' o ' o ’ o ’ o ’ o ' o ’ o ' o ’ o ' o ’ o ' o o o '

o NO o 'd - o "d- o in o c n o NO o CN oin CN in r- •"d- r - o o o r - CN NO I— I o in oo o •-11 o i— 11 o CN r H o o o o CN I— I 1—H

o ' o o ' o o ' o ' o ' o ' o ’ o ' o ' o ' o ' o ' o '

o m o CN o o o CN o CN o NO o in oN- c n o 00 00 m o CN d - CN d - o o CN i no o I— 1 o o o >—1 I ( o o o o I— I I— I oo ' o ' o c > o ’ o ’ o ’ o ’ o o ’ o ' o ’ o ' o o

o r- o NO o NO o m o 00 o NO o d - oCN y—( CN m CN c n CN CN 1-H CN CN CN CN CNo o o o o O o o o o o o o I— I oo ' o ' o ’ o ’ o ' o ’ o ' o ' o ' o ’ o ’ o ’ o ' o ' o '

00 r- 00 NO o o n - 00 d - 00 o 00 o 00 oo 00y—l o I—1 c - 1 - ^ ■'d- 1— t 1— 1 CN 1 - ^ 1— 1 t*h o 1— 1

o o o o o o o o o o o o I— I oo o ' o ’ o ’ o ’ o o ’ o ' o ’ o ’ o ’ o ’ o ' o ’ o '

NO ^ r NO c n NO r - NO CN NO r - NO NO NO r~" NOT— ( o T“ H in 1— 1 c n 1— t i— i 1— 1 1— 1 o T-h 00 T-H

o o o o o o o r— • o o o o o o oo ' o ’ o ' o ' o o ' o ’ o ' o ' o ’ o ' o ’ o ' o ’ o '

TJ- r- ■'d- ON "d- CN 'd - CN d " c n d " On d " ON d -1—H o in CN 1— t O i— i CN 1— 1 o 1 ( ON 1— 1

O o o O o O o o o o o o o oo o ' o o ’ o ' o ' o ' o ' o ’ o ’ o ’ o ' o ' o ' o '

CN , — 1 CN o CN ON CN NO CN 1— 1 CN CN CN CN CN-H 1—H 1—H CN T— On c n r -H ^ H 00 r H

O o O c n o o o O o o o o o o oc > o ' O o ' o ’ o ’ o ’ o ' o ’ o ' o ' o ' o ’ o ’ o '

o o O in o m o CN o CN o c n o NO of— I o 1— 1 CN 1— 1 CN 1— 1 On i— i o 1— 1 H 1— 1 r- r -H

o o o O o o o O o o o o o o oo ' o ' o ' o ' o ' o ' o ’ o ' o ' o ' c > o ' o ’ o ' o '

Pn Pn Pn-C 4 -> yG, 4-> J p 4 -* -3 4 -> -3 4-> JS 4-> -3 _4—> -34-> "o 4 -» o 4—* o 4-> o 4-» ’o 4-> " o 4 -* 'o 4->c x o c x c x CX JD CX o c x c x c x0) <L> CD <D CD CD © CD

Q ©> Q > Q ’flD

> Q CD

> Q > Q > Q 1 3

> Q

cd-t->cn(-X

CD

3>

3 T3 O cd1/5 — H

ed cn ©

CfHou.©

*1

©3 ,£ed

C/D

OnO n

«3

cd

cd3'O

ONm

cn

3

OOCNTT00

CNON»ri-d-

C/NNO

o m o c n C "c n o ON oo oo ' o ’1 o ’ o ’ o '

oo

c3

mCN

ocn

■'d"oo

o

ed3d3<d3T3

cdcnCNtd3TD

J=>cnCN"S3T3

<N3T3

edcnmtd3d3

r-in

moo'

cnrotd3TO

Velo

city

0.019

0.0

05

0.010

0.0

20

0.014

0.0

31

0.014

0.0

13

0.019

0.

024

duat

3sc

10 -0

.03

n/s

0.75

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

40

0.060

0.0

80

0.09

0

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

j

Velo

city

0.005

0.0

08

0.014

0.0

28

0.028

0.0

19

0.015

0.0

14

0.006

0.

006

duat

4sa

10 0.8

0 **

36.97

De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.100

0.2

00

0.300

0.

420

Velo

city

0.086

0.0

77

0.060

0.0

68

0.082

0.0

88

0.087

0.

131

0.162

0.

177 o c n o oo o CN o c n O CN o n r o n r o Ov

I— I T-H nT CN t—i nT I -H nT n r c n 0 0 o v/n o vo OVnT CN CN o c n o CN O n r I— I n r c n i—i o CN CNo ' o ’ o ’ o ' o ' o ' o ’ o ' o ' o ' o ’ o o ’ o ' o ' o ’o c n o vo o I— 1 o CN o v/n o c n o n r o Ovo o 0 0 c n CN c n v/n vn o ov o ^H CN o o r~-c n <N 1—1 O CN O i— i o c n o nT m i—| o CN CNo ' o ' o ' o ' o ’ o ' o ’ o ' o ' o ’ o ' o ' o ’ o ’ o ' o ’o VO o o o o 0 0 vo Ov o VO o n r o r_H o OVo o CN c n v/n v/n o n r CN c n o o o r - oo r ^ n r c nCN C N r c O T— 1 o I— I o o o CN I— I c n CN o o I— I CNo ' o ' o ’ o ' o ’ o ’ o ’ o ' o ' o ’ o ' o ' o ' o o ’ o ' o ’ o ’o 0 0 o OV o r -H o o c n c n o CN O v/n o c n o c no r - VO CN oo CN v/n cn CN T“H o v/n v/n r -H n r 1 -H nT voI—1 i— 1 o o o O o O O o I— I r -H ^H CN o o o r H

o ’ o o ’ o ' o ' o ' o ' o ' o ’ o ’ o ’ o ' o ’ o ’ o ' o ' o ' o ’o Ov o o o o nT o c n o c - o vo o c - o CN o o o<N i— i CN c n CN o CN r H CN o CN n r CN T— 1 CN I— I CN CNo i— 1 o O o o o O o o o o o I— I O o o ^Ho ' o o ' o o ' o ' o ' o ' o ' o ' o ' o ' o o ' o ' o ' o ' o ’oo r - 0 0 c n 0 0 r - 0 0 v/n o o n r 0 0 Ov 0 0 VO 0 0 c n 0 0 nTr*H nT r -H CN I -H I -H r -H CN I - H vo r -H r - I -H c n I - H 1-H T—H v/no O o O O o o o o o o o o t—h o o o I— I

o ' o ' o ' o ' o ’ o ’ o ’ o ' o ' o ' o ' o ' o ’ o ' o ' o ’ o ' o 'VO CN VO n r VO un vo 0 0 vo o vo CN vo o VO o vo or -H i— i i— i CN t- h o I—1 I -H 1— 1 T-H I— I r - 1 -H 0 0 I -H I - H T“ H CNo i—1 o o o o o o o o o o o o o o o r -H

o ’ o ' o ' o ' o ' o o o ' o ’ o ’ o c l o ’ o o ’ o ’ o ' o 'n r VO n r nT n r n r nT cn n r r - n r n r nT nT Ov n r oI-H Ov I— 1 CN f—H i—i I-H I^H t- h o I -H 0 0 I -H o I -H o 1 -H nTo O o o o o o O o o o o o r -H o o oo ' o ’ o ' o ' o ' o ' o ' o ' o ' o ' o ' o ’ o ’ o ’ o ' o ' o ' o 'CN Ov CN r*H CN Ov CN c n CN vo CN n r CN VO CN vo CN t - Hr -H VO I -H m ^ H t- h I -H ^ H ^ H o I -H r ~ - I -H OO I -H o r*H CNo O o o O o o O o o o o o o o o oo ' o ’ o ’ o ' o ' o ' o ' o ’ o ’ o ' o ' o ' o ' o ' o ’ o ’ o ' o ’o c n o oo o n r o oo o CN o n r o c n o r ^ o 0 0t- h vo I -H n r I -H o I -H o I -H o r*H nT r -H O r -H o p H r -H

o o o ^ H o o o o o o o o o o o o t- h

o ' o ' o ' o ’ o ' o ’ o ' o ' o ' o ' o ’ o ’ o ’ o ’ o ' o ’ o ' o '>% >>

-C p p -C Jp •*-» rP ,P H—> , p H—< rP •/-> rP •£2o -4-» o •*-» "o H-* o 'o H—* o H—» o H—* o + -» o

c x j d c x o c x c x c x j o c x o c x o c x o c x J OCD CD a> CD CD CD CD CD 0)Q > Q

*o3> Q

'aS> Q

’a3> Q

CD

> Q > Q’a3> Q

*a3> Q >

VOOOr-VO

OO00

cnOv

VOOOo ’(N

oo ’

OvVO

VOO n

OOVO

0000

00 ocn

cdP

TD

cdp

TD

cd

cdP

TD

cdP

TD

OO

00

O cn VOtdP

TD

OVO

00Ocn<N

voOV

OOo

oo'

CNcncnoo

voOV

cdcdVOcdPTD

ed

edP

TD

cdP

TD

edc/500cdP

TD

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

jdu

at9s

a 10

0.19

n/s

3.16

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

70

0.150

0.2

30

0.30

0Ve

locit

y 0.0

13

0.031

0.0

21

0.02

1 0.0

30

0.035

0.0

37

0.038

0.

031

0.02

9 du

at9s

b 10

0.95

***

174.3

0 De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.050

0.1

00

0.150

0.

200 o o O o in

oo SO CN 00 nrcn 1—*1 nT o CNo ' o ’ o ’ o ’ o ’"sf o o o o m oooc CN Os cn cn CO CNCN 1 cn >—i1 nr O CNo ’ o ’ o ’ o ’ o ’ o ’ o 'Oc o O o o O nrin oo T—l OC o m OsCM o cn o nr o I—Io ' o ’ o ’ o ’ o o ’ onr o r - o r - m COOc nr Os in CN cn 001—!1 o CN o cn Oo ' o ’ o ’ o ’ o ’ o ’ o ’00 o CO o cn O mnr <N CO CN i—i CN CO1—11 o CN o cn oo ‘ o ’ o ’ o o ’ o ’ c>cn 00 cn 00 CN 00 00nr t-h i—i 1—1 O H CO

o CN o cn o T—1o ' o ’ o ’ o ’ o ’ o ’ o ’O n CD o CO o CD CN<N cn r-H CO 1-H r -T-H o CN o CN O *—•

o ’ o ’ o o ’ o ’ o ’ o ’in nr nf nr in nro 1—H T— 1 r ■■ H r-H ^H mT—' o CN o cn O 1-H

o ’ o ’ o ’ o ’ o ’ o ’ o ’o CN cn CN o CN nro i—H o i—i r - 1—H nrI—I1 o CN o CN O I—Io ’ o ’ o ’ o ’ o ’ o ’ o ’00 o CO o CN o OSOs i—1 Os H CO I— I nro o i—■ o CN o I—*o ’ o ’ o ’ o ’ o ’ o ’ o ’Sn >>

-C X X -+—*7 3 7 3 H-* * 5 H-> oo cx o cx j d CX j d

<D 0> <D7 3

> Q7 3

>Q

7 3

>Q

<D

>

00 cn CN\ ] m nTcn CN 00XI in 00

***

**

***

nT r -Os oo Oco ’ c> o ’

O Os o

cd x > ou s u s u sO O oi- 1— 4 r-H

cd cd cdP 3 3

T3

o CO o oo oo nT o oCN cn t-~ nT CN r - H CN occn CN o CN o I— 1 CN cno ’ o ’ o o ’ o ’ o ’ o ’ o ’o cn o m co m o CNn r m CO CO CN CN r - COCN CN o CN o I— I 1— I CNo ’ o ’ o ' o ’ o ' o ’ o ’ o ’o cn m m nT cn o oCO r - n r i n CN CN r — 1 e ' r - H CN o CN o I— I I— I eno ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’o cn o oo CN I— 1 o o00 CN cn CN CN cn CO n ro CN o CN o I— I o n ro ’ o ’ o o ’ o ’ o ’ o ' o ’o co o o o cn o oCN n r CN Oc CN o CN oo I— 1 o T -h o l—H o n ro ’ o ’ o ’ o o ' o ’ o ' o ’00 f - oo OC 00 o 00 CO1 -H n r I— I o I— I cn I— I n ro I— I o cn o I— I o I— I

o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’CO CO 00 CD o CO nT1 -H r - ^ H r - I— I Oc r - H oo r - H o I— I o o o CNo ’ o ’ o ' o ' o ’ o ' o ’ o ’n r CO nT Oc n r CN n r 00I— I nT I— I 00 1 -H o 1— 1 ooo o I— I1 o 1 -H o oo ’ o ’ o ' o ' o ’ o ' o ’ o ’CN CN CN CO CN CN CN1— 1 CN I— I r - I— I o r - H Oco I— I o I— I o I— I o I— I

o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’o ,— I o m o oo o Ocr - H cn 1— 1 CO 1 - ^ o I— H CNo I— 1 o r " H o I— I o CNo ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’

> 1 > 1 > >-C H—> X +-» -C X H -*

H—1 73 o o 73c x o c x j d c x JD c x JD<L> (L> <1>

Q73> Q

7 3> Q

73> Q

7 3>

Oc<25 m Oc CO

cn CN r -nTCO m cn n r

* u s u s**

m

r - cn o Oc00 cn CN CNo ’ o ’ o ' o ’

o o o or - H I—H

cd X o edu s u s u s u s

1 i I— I I— I CNN—> H-* H-H H-*x > pO X x3 3 3 3

n3 T 3 T 3 T 3 dubt

2sb

10 0.7

0 *

22.2

1 De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.090

0.1

80

0.270

0.

350

Velo

city

0.183

0.0

59

0.170

0.0

88

0.090

0.1

49

0.197

0.2

26

0.310

0.

310

dubt

2sc

10 0.2

7 n/s

4.2

8 De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.024

0.0

28

0.032

0.

038

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

j

Velo

city

0.109

0.1

39

0.157

0.1

63

0.162

0.1

62

0.166

0.1

46

0.176

0.

152

dubt

3sa

10 0.4

4 n/s

1.4

1 De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.050

0.0

90

0.130

0.

180

Velo

city

0.180

0.2

22

0.215

0.2

08

0.214

0.2

18

0.172

0.1

85

0.219

0.

289 o o oo o o

CN c n CN i f VO<N 1 f o o CNO o ' o ' o ’ o ’O o VO i f oVO 0 0 (N i f Ov'—1 c n O o i— i

o ' o ' o ’ o ' o ’o CN " s f Ov o! H r - CN m c nI H (N O o >—•O o ’ o o ’ o ’o ON CN CN oVO I f CN i f t " -O CN o o oo ' o ' o ’ o ’ o ’o _ o o o<N <N CN o CNo CN o m oo o ' o ' o ' o '0 0 OV 00 r - 00t- h I— I r -H i f T-h

o CN o o oo ' o ' o ' o ' o 'VO CN VO VO VOt—h <N i— i i f I -H

o CN o O oo ' o ' o ' o ' o 'i f Ov i f O i f! H r - i— i I -H

O t—h o o oO o ' o ’ o ' o 'CN c n CN 0 0 CNt—h in i— i i f I - H

o o O oo ' o ' o ' o ' o 'o CN o VO oT—H i f I— I CN I— I

o i— 1 o O oo ' o ' o ' o ' o ’

> >

x z xz - CH~» o o -!-*c x o c x j o c x<D CD CD

Q1 3

> Qo

> Q

o O o m i ■ i

oo OO OV c n Ovi f I— 1 i f o Oo ’ o ’ o ' o ’ Oo o O CN _

r - i f 0 0 c n c ni f t- h I f o i— i

o o ' o ’ o ’ o ’vo o o oo 0 0Ov OV vo CN 0 0CN o i f O oO o ' o ' o ’ o ’i f o o i f ooe n <n in CN ooCN o c n o oO o ’ o ' o ’ o ’r - o o o I— ,r H CN f " CN Ov

o c n o oo o ' o ' o oCN 0 0 o 0 0 T-h

r - I -H ' i j - I -H oor H o c n o oo ' o ' o ' o ' o 'r - vo o vo T—H

m ^ H c n I -H oo*—• o c n o oo o ' o ’ o ’ o 'CN i f o i f r -OO I -H r - i— i OvI— 1 o e n o Oo ’ o ’ o ’ o ’ o ’r -H CN Ov CN r -c n r -H ' l i r -H OvT i o e n o Oo ’ o o ’ o ’ o 'CN o i f o ' i i -r -H T-H Ov t- h voi— l1 o CN o oo ’ o ' o ’ o ’ o ’

-*->X Z

H-> xz H-*

o H-> o H-> Bo c x JD CX o<L) 0 )CD

> Q’B> Q

CD

>

o o o o VOo CN o Ov c nCN c n CN in oo ’ o ’ o ' o ' o 'o o o o CNm 0 0 in i f c n

CN 1—' VO oo o ’ o ’ o ' oo r - o o 00o e n o o CNI—1 1 >—1 m o(O o ’ o ' o ’ o 'o 00 o o I fm CN m c n CNo o o i f oo ’ o ' o ’ o ’ o ’o vo o o oCN i—i CN m CNO o o c n oo o O o ' o0 0 VO 0 0 c n ooT-h T-h t- h CN ^HO o o c n oo ' o ’ o ' o ' o 'VO VO VO Ov vot- h T -H t—h o t- h

o o o c n oo ’ o ’ c i o ’ o 'I f o O I frH ^ H t—H i f 1—H

o o o c n oo ' o ' o ' o ' o 'CN r - CN O CNI - H o r - H r - Ho o o CN Oo ' o ' o ’ o ' o 'o vo o o oI—1 o I— I m I - H

o o o CN oo ’ o ’ o ' o ' o '

xz H-* rC xzH-> o H-> o ■Hcx JO cx o cxa> CD CD

Q > Q(D

> Q Velo

city

0.009

0.0

17

0.018

0.0

18

0.020

0.0

16

0.02

1 0.0

24

0.018

0.

035

dubt

6sa

10 0.9

7 **

* 29

3.93

De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.050

0.1

00

0.150

0.

190

Velo

city

0.049

0.0

71

0.082

0.0

95

0.120

0.1

87

0.260

0.5

10

0.600

0.

590

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

jdu

bt6s

b 10

0.73

* 25

.23

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

50

0.080

0.1

10

0.14

0Ve

locit

y 0.6

00

0.620

0.6

10

0.650

0.6

40

0.650

0.6

70

0.730

0.6

80

0.69

0 du

bt6s

c 10

0.73

* 25

.23

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

35

0.050

0.0

65

0.08

0 o o C" o oo o o o o o o o o o o o o oO s O s CO n r s o n r c - r H O s CO r~~ O s 00 o c n 1-H c n O sc n o CM CM o CM nT n r '—I oo o in 1 H 00 I -H voO o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’O m 00 o 00 o o o o o o o o o o o o oe' r - r - 00 O s 00 00 00 s o r-^ m c - O s 00 O s 00 r - ine n o o 1—< o s o o nr 1—* 00 o m o c - o s o I -H

O o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ' o ’O o 00 o CM o o o o o o o o o o o o o>—i s o O s CM O s CM 00 s o CM oo O s m o vo CM vo 00 oc n o o 1 o I—11 VO o CO o C " o vo o r - o voo ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ’o I/O in O O s o o o O s o o m o m o o o oCM c o S O S O 00 s o CM n r 00 m oo CO S O c n O S n r c n mCO O O o o m o CO o s o o m o ■n- o VO oO o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o o ’ o ’ o ’CM o CM O o o o o o o o o o o o o o oO CM r - CM 00 CM CO CM C " CM r - CM nT CM c n CM O s CMCO o •—11 O o o n r o CM o m o m o n r o m Oo ’ o o ' c> o ’ o ' o o ’ o ’ o ’ o o ’ o o ' o ’ o ’ o ’ o 'm 00 cm 00 o 00 o oo CM 00 o 00 o 00 o o o o ooO s I— I CM T < XT I -H o r*H I -H 1 1 O s I -H in I -H nT r-H o I - H<N o »—• o o o n r o CM o n r o m o nr o vo oo ’ o ' o ’ o ' o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o o ’ o ’ o ’CM s o m S O O s s o o S O m vo o vo o VO o VO o vot" - I -H m T—< n r I -H n r I -H CM l - H I-H I -H 00 I -H CM 1- I -H I -H

CM o i—1 O o o CO o o nr o m o n r o s o oo ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’o n r O S n r O s n r o n r in n r o n r o n r o n r o n ro i—i CM I— 1 00 I -H n r CM I - H O s I -H vo T-H nT I -H o I -H

CO o o o o CO o o o CO o m o n r o i n oo ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ' o ’o CM O s CM CO CM o CM r - CM nT CM o CM o CM o CMO s O I -H O s i n I -H o r—* r—4 n r I -H r » r—H 00 C“HCM o i—1 o o o CO o o o CM o m o c n o m oo ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ’O S o O s o so o o o r - o o o o o n r o o oOO I -H so I -H CO I -H o I - H O s 1 -H 00 I -H m I -H n r I - HCM o o o o o CO o o o CM o n r o CM o m oo ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’

Sn So So Sn Sn_-t—> -C JC _-t—> .■£3 -C ."£3 ,£3o o -»-* o o ■*-> o o +-* o +-• "c3 QJ O CX JD cx jd cx o cx o cx o cx jd cx jd cx o cx <D <D <L> <D £> &>d>> Q 13

> Q 13> Q 13

> Q 1 3> Q 13

> Q 13> Q 13

> Q 13> Q

Ii

o’■fj cr>•cdc/2

Mh

<DCl>

P<

C/2“e 'O ’O cd«-i crcd C/2 "SS o

O h

£

O

>noo ' o ’ o 'oom

OCNO

VOr -o

o o ’ o<nr ^CN

oor -H

ovoi no

o o ’ o ’mc nCN

voo 0 0o

o ’ o ' o 'T *

mCN

■d"

ovoi no

o ' o ’ o ’ooCN

CN

oONi no

o ' o ’ oVOCNCN

oo

r -■d-o

o ' o ’ o '-4-*

GJO

Dep

th

>2■*-»

'oo13>

(L>>

'd-vqinVO

***

VO00o ’

o

x>C/2ox>3T 3

ooCNONr~"t i

oNO CNr-"oo»N<N

00CNCNo00CN

ONCN^ H

oCNr-H

ood" o00CNCNCNCN

oNOr"o 00 1—H

o ' o ’ o ’ o ’ o ' o ' o ’ o ’ o ’ o o ’ o ’ o ’ o ’oNO OnIN oCN

r -H

r~-o ocnCN

CNCNoooCN

cnr -H

oCN d-CN ooCNd-CNCN

ooNOCNoo or"00

o o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’o00o

ONCNr*H

o00o NOoo•N o00o

ooCNCNCN oooo

od- ooCNCNCNr -H

ood"00oCN

oor-o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’od"o

oCN od"oCNo

ocno

CNi—1r—1oo CNooo

od"oCNCN oo CNd" ooCN

NOCN ooNOo o o ’ o ’ o ' o ' o ’ o ’ o o ’ o ' o ' o ' o ’ o ’o(No

00CNhhoCNo CNo

oCNo00CN oCNo

t- h

r"ooCNo

OnT-H

oCNocno oCNo

NOCNoooCN

o o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ' o ' o ’ o ’ o ’ o ’00o

00CNI i00r -H

or-CNo

00o

NOCN1—-t00o

ooNOo00o

r--o 00o

CNONoNOCNo

CNo

CNd"oo ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’NOrHo

00d"r -H

NOo

CNCNoNOo

NOf “H

NOr -H

o NOoNOT“ H

o00 NO

oo00o

CNCNor-00o

od-oo ’ o ’ o ’ o ’ o ' o ’ o ' o ’ o ’ o ’ o ' o ' o ’ o ’ o ’d-o

r-CN! 4d"r -H

oOnCNo

d"r -H

od"d" d-

oCNNOo

TfHo

1 Hd"fHo

CNd"oooo

r-cno

oCNoo ’ o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ' o o ’ o ’ o ’ o ’CNo

INd" CNo

OnCNoCNo

Ono CNo

od“oCNo Ono

CNr -H

oCNNOo

d"o

r-CNooCNo

o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ohHo

o oo

r"d"ooo ooo

oT—Ho

oCNooo

Onr-ooo

d"CNooo d*o

oo

o ’ o ’ o ’ o ’ o ’ o ’ o ' o ' o ’ o ' o ’ o ' o ' o ’ o ’

Dep

th ’oo<D> D

epth

>>H-»‘6oI d> D

epth ‘oo

<D> D

epth H-H

oo1 3>

Dep

th

>3'ooCD> D

epth +-*

o1 3> D

epth o13> D

epth

I— I ON NOoo I— 1 I— 1

CN 0 0 o '

On o r-- ooCN CN CN NOCN r-i CN NOr-~ d- d- ON

ON

CNC/3

s ^ = H

cn d" ON CN CN I—1 I—1CN d" oo oo oo ON r—lo ’ o ’ 1 o ' o ’ o ’ o ’ o1

NO cn

© £

K ■£►—( -*-* o cdC/3<NH-»cd-t-«<d

.©C/3CNtdI d

oC/3<NcdI d

cdC/3

c n-*->cd-*-»cd

C/3c n

tdId

oC/3

c n-+->cd-♦-»CD

cnd-tdtd

Velo

city

0.071

0.0

91

0.145

0.0

62

0.147

0.0

25

0.053

0.0

19

0.34

3 et

at4s

c 4

0.31

5 5.1

1 De

pth

0.010

0.0

15

0.020

0.0

25

0.030

0.0

35

0.200

0.4

00

0.600

0.

650

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

j

Velo

city

0.138

0.1

96

0.206

0.2

95

0.276

0.2

80

0.274

0.2

77

0.266

0.

297

etat

5sa

10 0.8

**

37.33

De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.050

0.2

50

0.450

0.

550

Velo

city

0.012

0.0

57

0.044

0.0

65

0.049

0.0

90

0.202

0.1

60

0.213

0.

223 omm

oNO 000<NNO00 00CN

cnNO00CNin

CNNO 00cn

0cn 0n -r-~

NOO OOcn mOOCNin

OnOr—sO00N"

CN0

o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ O O o ’ o ’ o ’ 0 o ’omTf-

oom OCNCN ON Om 0N-00in'sf

NOmOOinCN

r-CN 00NOONr-0

OOCNNOr0

0mn-CNO 00cn

NOOOo' o' o' 0 o ’ o ’ o ’ o ’ o ’ o ’ O o ’ o ’ o ’ o ’ o ’ o ’ o ’omCN

oOnOONO1“ H cn 00 NO■rr0

OOcn

r~~000inr“H

ininOOO mNO0

0in ON0

00cn 0r--l

OOCNCNCNO

o' o ’ o' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’omo

inr-oOO 000 0mO

NOr0Oint i

mCNO0mO

cnO

00CNcn000

OOr" i r0

0mr H

OnNOOOO 0cnO

o' o ’ o' o ’ O o ’ o ’ 0 o ’ o ’ O o ’ o ’ o ’ 0 ’ o ’ O o ’oCNo

OnO

O<NO00

OCNOt"-mO

OCNONOOO

0CNOinCNO

ONOO■'sf-ONO

0CNOCNNOO

mCN0r"-m0

mCNONOCNO

o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ O’ O o ’ o ’ o ’ o ’ 0 o ’ o ’ o ’00o

cncno000 ■"sf0

OOO

cncnO

000

00OO000

cnr—HO

0inONOr"0

00HO

r-n-0CNCNO

r—•cn0CNCNO

mCNOo' o' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’NOo

NOinoNO0

00

NO0

CNtTO

NOO

cnO

NO^H0

NO0000

CN000NOr IO

cnr-0OnO

ONm0ON0

r-CN0o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’

o mo 000

Tl-1—HO CN0

N-0

OCNO 0NO00

0cnOmNOO rHO

On

ONOO

00CNONO0 CNO

o ’ o ’ 0 o ’ o ’ o ’ o ’ O o ’ 0 o ’ o ’ O o ’ o ’ o ’ 0 o ’CNo

oinOCNO

incnOCNO

0000CN0

NOcnOCNp* HO

00000CNO

CNinOCNO

ONCNOcn0

’'t0

cn0

H

Oo ’ o' o ’ o ’ O o ’ o ’ o ’ o ’ o ’ o ’ O o ’ o ’ o ’ o ’ o ’ o ’oo

CNTTO00

Orf-OOO

000

00 O

0r-H0

cncnO00

NONOOOr-H0

OnNOO00

0r -0

00

OO

o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ O o ’ o ’ o ’ o ’

Dep

th -t—*‘5O

> Dep

th -t—•’ 0JO Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

thV

eloc

ityD

epth H->’O

JO<D

> Dep

th

>>’ 0O'to> D

epth ’O

O<D

> Dep

th +->OO

>

mNOr~"

cnooON

OnNOCN

CNr r-CN

ONcn

ooNO cncn

CN

t/3 C/5

’sT00 cnON

r-(N inr <No'

ino ON

NO«>o

NO

c / 3 o n c / 3 c « c « c n t / 5 i / i t / 5m m v o N O N o r ^ r ^ - o o o oa j ( D ( U < D a )

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

j

etat

8sc

5 0.2

7 n/s

4.3

7 De

pth

0.010

0.0

15

0.020

0.0

25

0.030

0.0

35

0.150

0.3

00

0.450

0.

580

Velo

city

0.034

0.0

49

0.038

0.0

67

0.069

0.0

60

0.058

0.0

59

0.082

0.

060

etat

9sc

6 -0

.1 n/s

0.1

5 De

pth

0.010

0.0

12

0.014

0.0

16

0.180

0.0

20

0.040

0.0

80

0.012

0.

016 o

00o

o00cn

oC"VO

oocn

VOCN

on rCN

r-"m

ooo

cnom

oCNCN

oot"-

OVOcn

voi nCN

00CN

o ' o ' o ' o ' o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o '

nTC"o

oOcn

ooor -

ooCN

oOscn

o00

oovo

oi n1—H

CNr -n r

ooo

oVOm

onTCN

omcn

0i n^H

o ' o ' o ' o ' o ' o ’ o ' o ' o ' o ' o ' o ' o ’ o ’

I-"mo

OOCN

or "nT

oi n

ovocn

oo

oH

oo

r- oo

cnCN

ovor -H

00cncn

00

o ' o ' o ' o ' o ' o ' o ’ o ' o o ’ o ’ o ' o ' o ’

n rSOO

oo

on rcn

oo

cnvocn

omo

oOsVO

oi no

cnoo1—H

omo

VOO

oooo

CN

CN

0m0

o ' o ' o ' o ' o ' o ’ o ' o ' o ' o ' o ’ o ' o ' o '

cnr "o

oCNo

VOmcn

ocno

oOscn

oCNo

VOmcn

oCNo

oOscn

oCNo

Oscn

oCNo

OnTo

oCNo

cnn r

0CN0

o ' o ' o ’ o ' o ' o ' o ' o ’ o ' o ' o ’ o ’ o ’ o o ’ o '

CNr ô

00

omnTcn

VOCNO

nT00cn

00

oi nn rcn

00

oom

00

or-m

00

oCN 00

ovocnr -H

001-H0

o ' o ' o ' o ' o ' o ' o ' o ' o ' o ' o ' o ' o ' o ' o ' o ’

nTVOO

vo

ocncncn

CNCNO

OsCNCN

voT—1o

cncncn

vo

oor^-

vo

oCNn r

VOr -H

oVOr- vo

oOscnr H

vo

0

o ' o ' o ' o ' o ’ o ' o ' o ' o o ’ o ' o ' o ’ o ’ o ’ o '

nTnTO

nT

on rJ—1cn

00 1—H

oO S

n rr -H

onT

cn

n r1-Ho

ooovo

nT

ocn

n r

o00OSo

nTr -H

oOS

CN

n r

0

C> o o ' o ' o o ' o ' o ' o ' o ’ o ’ o ' o o o ' c>

00n ro

CN

or -H

cn

n r

on f

cn

CNr —H

O cn

CN

oon rvo

CN

or- CN

ovoo

CN

omn r

CN

0

o ' o ' o ' o ' o ' o ' o ’ o ' o ' o ' o o ' o ' o ' o ' o ’

i nVOO

o?“H

onTvo

oo

nTCN

O

O

nTvor*H

oo

t"-00CN

oo

VOOsO

oo

cncnr -H

oo

Os^ H

o0

0

o ' o ' o ' o ' o ' o ’ o ' o ' o ' o ' o ' o ' o ' o ' o ' 0

4 -*

'5J O

> De

pth

Ve

loc

ity

De

pth

> >

' 3o

De

pth

-t—■* 3

o1 3

> De

pth ' 3

o1 3

>

De

pth

> >

* 3o

1 3

> De

pth ' 3

JO<D

>

De

pth

S n-t—*

■ 30

1 3

>

De

pth

inin CN

nTC N

cncn

invo VO

nToO s

COOsC"

oO S

lb in

vo

o1/3

o

X>O

oc/3

CN-4->

<L>

cdC/3

CO-4-Jx>-t-*<L>

x>C/3

CO+-*x>-4-*<L>

Velo

city

0.490

0.6

30

0.420

0.6

20

0.199

0.5

50

0.550

0.5

90

0.760

0.

860

etbt

3sc

10 0.6

7 *

19.70

De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.070

0.1

20

0.170

0.

220

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

jVe

locit

y 0.0

71

0.266

0.2

03

0.094

0.2

86

0.287

0.5

70

0.660

1.3

60

0.63

0etb

t4sb

10

0.51

5 10

.47

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

40

0.080

0.1

60

0.24

0Ve

locit

y 0.1

91

0.264

0.2

25

0.266

0.2

30

0.273

0.2

75

0.29

1 0.2

77

0.30

7 ooc n

or rm

o00

c n oVO1—

oVO

oCN

oTj-vo

or -o

ON^ rCN

ovoCN

oc -c n

or -o

C"oom

oo

m

c n

oCN

or~~in

o ’ o ’ o ’ o o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’

ooCN

Ooovo

oin

CNOin

oCN

oO nVO

oo

oooVO

omo

oVOin

ooCN

oCNin

omo

c n

i n

inc nO

c nO nCN

oONo

omin

o ’ o o ’ o ' o o o ' c> c> o ’ o ' o ’ cd o ’ c d o ’ c d o ’

oo

oovo

oo

oo

T“^

o00o

oTl-vo

o00o

oc -vo

on -o

-ri-o o

oot t

o-'Cj-o

voc n

oc no

OOc nCN

ovoo

oc nm

o ' o ' o ’ o ' o ’ o ’ o ’ o ’ o ' o o ' o o ’ o ’ o ’ o ’ o ’ o ’

omo

oinin

oino

00 oo

oCN00

oTl-o

oC"vo

oc nO

oooCN

oooo

omTt-

oc no

oc nc n

mCNo

m

CN

oc no

o-tTin

o ’ o ' o ' o ' o ’ o ' o ’ o ' o ' o ’ o ' o ' o ’ o ’ o ’ o ’ o ’ o ’

oCNo

o00m

oCNo

c nCN

oCNo

om00

oCNo

oONm

oCNo

omc -

oCNo

t TmCN

oCNo

c nOc n

oCNo

CNCNCN

oCNO

oVOm

o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ c d o c d o ’ o ’ o ’ o ’ o ’ c d c d

00

oovoi n

00 r"H

ooooo

00

oomr "

00 r —<

oof— i

VO

00

oo^ rCN

00

oo-Tie n

oo

ovoc nc n

00

ooCN

00

ooc nm

o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ <d o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o '

voo

oom

VO

Oc n

VO

ooCOr -

voo

ovovo

voo

r -oc n

VO

oor -c n

VO

oc nCN

voo

T fmCN

voo

om

o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’

T f

ooc nc n

T f

oo oVOo

T f

ooOOr -

Tj-

oooc -

-n-

or

CNT f

ovoCNc n

T f

oo00o

Tj-

o-TfT tCN

-Tf

oomin

o ’ o ’ o ’ o o o ’ o ’ o o ’ c> o ' o ’ o ’ o ’ o o ’ o ’ o ’

CN

ooON^ r

CN

oCNOOo

CN

or -VOVO

CN

oovovo

CN

ooooCN

CN

omc nc n

CN

oc nCN

CN

ooCNCN

CN1 <

oom

o ’ o ' o ’ o ' o ' o ’ o ’ o ’ o ' o ’ o ' o ’ o ’ o ’ o ’ o ' o ’ o ’

oo

ot "" 3 -

oo

i nc no

oo

OCNVO

oo

ocnr -

oo

COooCN

oo

Oi nCN

oo

oi nt t

oo

CNc nr—l

oo

oc nm

o ’ o ’ o ’ o ’ o ' o ’ o ’ o ’ o ’ o ’ o ' o ' o ’ o o ’ o ’ o ’ o ’

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

thV

eloc

ityD

epth

Vel

ocity

Dep

th -t-J‘ 5o Dep

th

>>

oo'<5> D

epth

Vel

ocity

Dep

th ’o

> Dep

th>> -*—*oo13>

'aCO

'a 'a a

00c n

voCN

CNO O nO i nON c n

m CNO

VO VO VO VO

0 c d j D 0 « j , 0 > 0 « j 0^ j - u ^ i r N v n v o v o v o r ^ r ^

N—» -f-> +->( u a > ( U ( D

Sam

ple

Num

ber

of Pe

arso

n's

P va

lue

F st

atist

ic

poin

ts Rs

q ad

jet

bt8s

a 10

0.48

5 9.5

4 De

pth

0.010

0.0

12

0.014

0.0

16

0.018

0.0

20

0.040

0.0

80

0.012

0.

016

Velo

city

0.083

0.1

03

0.042

0.0

42

0.073

0.1

35

0.137

0.2

08

0.10

1 0.

017

etbt8

sb

10 -0

.04

n/s

0.61

Dept

h 0.0

10

0.012

0.0

14

0.016

0.0

18

0.020

0.0

40

0.060

0.0

80

0.10

0 ^r o (N o Ov o o o o o 00 o c ncn o o t-" o Ov CN 00 oo Ov o cn o c ncn 1—1 cn I—1 cn r ^ r-H ov I—1 cn c n oo ’ o ’ o o ' o ’ o ' o ' o ' o ' o ' o ' o '

o o o o o o Ov o cn o CNOv VT) r - oo r r 00 CN cn 00 o voc n '—1 VO o cn CN 00 1-H 1-H o CN CN oo ' o ’ o ’ o ' o ’ o ' o ' o ' o ' o ' o ’ o ’ o '

<N o _ o o o o r - o oo o voOv o CN vo cn CN o o CN VO o vo cn^r I—* CN o CN i—1 VO 1—' T—H o c n 1-H oo o ’ o ' o ' o ' o ' o ' o ' o ' o ’ o ' o ' o '

w o OO o o cn o cn o vo o vor - irv oo 00 00 OV vo CN CN 00 Ov

o I—11 o I—1! o cn o cn o CN o oo o ' o ' o ’ o ' o ’ o ' o ' o ' o ' o ’ o ' o ’

r - o vo o vo o o o o o w o cnOv (N Ov CN o CN r - CN o CN CN CN CN^r o I— 1 o o CN o cn o c n o oo o ' o ' o ’ o o ’ o ' o ' o ' o ’ o ' o ’ o ’

00 00 cn 00 Ov 00 00 cn 00 00Ov 1—1 Ov 1—1 cn 1-H Ov CN 1-H vo 1-H CN^r o I—11 o 4—11 o i—' o o CN o oo o ’ o ' o ' o ' o ’ o ’ o ' o ’ o ' o ' o ' o '

VO vo r" vo oo vo Ov vo "3- VO ov vor - 1—1 Ov 1-H un 1-H cn r-H CN 1-H Ov 1-H cn’s f o 1 o i—' o I—1 o CN o CN o oo ' o ’ o ' o ' o ’ o ' o ' o ’ o ' o ’ o ' o ' o ’

cn c n c n N" o 'nT CN c n Tj- o<N 1—1 1-H 1-H 1-H t" ' 1-H 00 1—1 c n 1-H vo

o »—11 o 1-H o 1—' o 1 -H o CN o oo ' o ’ o ’ o ’ o ' o ' o ' o ’ o ' o ' o ’ o ' o ’

VO <N o CN r - CN o CN c n CN c n CNi— CO 1-H Ov r— l CN 1-H 00 1 -H c n 1 -H r H

cn o i—' o o o o 1 -H o CN o oo ' o ' o ’ o ’ o ’ o ’ o ' o ’ o ' o ’ o ' o ' o ’

m o CN o cn o CN o Ov o Ov o cnvo 1— 1 ■*fr 1-H vo 1-H oo 1 -H cn 1— 1 ov 1-H c nr -H o o o o o o o CN o oo ' o ' o ’ o ' H o ’ o ' o ’ o ' o ' o ' o ' o '

.ti x:oo13>

CX<L>Q

ojo13>

x: .tz x .tz x .ts x .ti x .t:cxa>Q

oo1 3>

cx<L>Q

o13>

cx<DQ

oo

1 3>

cx<L>Q

ooa>>

cx&>Q

Qo1 3>

<Ni/"> Oo' ooIT) c n

vo voOCN

t/3

cnO

<N

Or

Ovc nCN

Ov VO vo

ocn00-t-*X(U

xcnOvXa>

ocnON+->X-4-*a>

cnO

-4-»a>

xicnO

X>+-»<L>

OC/2o

Xa>

227

Appendix IV

Appendix IV: TWINSPAN ordered species by samples tables

Introduction

The results of the TWINSPAN analyses used in chapters 2 and 3, is presented here as order

species by samples tables. Taxon (OTU) codes are used throughout. The full taxon names are

given in table 2-5,chapter 2. In each matrix a and b refer to the deep and shallow sections

respectively. The last column and last row give the cuts produced by TWINSPAN. These are

represented as ones and zeros, the line o f binary digits closest to the rest of the matrix represents

the first cut. The numbers in the matrix are pseudospecies counts. Pseudospecies are a

representation of the species abundance in simplified form, (Jongman 1988) .

For all sites TWINSPAN failed to make a clear destination between samples from deep and

shallow sections.

228

Appendix IV

Table IV-1, River Etive data. The first major split is identified by using bold text to separate the two groups.

a b b b b a a a b b b b a a a a a b a b a a a b b b b b b b b a b a a b b b b b b b b a a b b b b b b bI-'P'g 3 --------------1 2 2 1 1 - - 1 - 2 2 2 3 - 1 - - 2 1 2 - 2 1 1 ---------- _ -------0Hept 2 4 - - ................................... - -------0Pebi . 1 ------------------------------------- -------0Hydr ---------- 1 1 2 1 1 2 4 4 4 3 3 4 3 4 - - - 2 4 - - 4 - 1 - 4 - 1 - - 2 ...................... - - - - . . . - -------1Oxyt - - 1 -----------------4 ------------- - 3 1 1 -------1 . . -------1Pofl 1 - 1 1 - - - 1 1 - - - - 1 - 1 - - l . . . . 2 - - - 2 . . . . l . -------1Rhse ................................... 1 ------------- . . . ] . l . . -------1Dyti - - ...................... ..................... I 1 - -------1Para . - 1 -------1 1 - - 1 ---------- . I . ] ---------- 1 ------- -------1Nilo ----------------------- 1 ---------------- -------1Macro . . i --------------------------------- -------1Thicn - 1 1 -------1Tany -------------------- 2 -------------------- - - - 2 - - 1 1 ---------- 11 - 1 1 - -------1Heter - - 1 - ........................................ -------1Pico - - 1 ............................................ ..................2 - - -------10Calu ...........................1 -------11Fxdy ---------- 1 ---------- -------1 - . -------11Siar .................. I ' -------11lemo 1 - l l ------- 1 - 1 -------11Pemi - ..................2 - -------11I.ob scui - 1 - - -------11Ortho - - 1 .................. -------11Virga 1 ------- -------11Diam ------- --- | ------- 1 ------- -------11Genus Ii - 2 ...................... 2 .................. -------11Cari 1 4 ................................................ ---------- 2 - - 1 1 - - 1 1 - - - 1 -------1Olig 1 1 ................................... 1 - 1 2 - - - - - - 1 2 1 - 2 - - I - - . 4 3 - 4 2 - - - 1 - - - - - 1 - - - - - - 1Barh ................................... 1 ----------2 - - 1 - - - 1 1 - - 2 2 11 1 2 - 3 - - - 2 3 2 2 2 - - 2 - - . . . - - - - 1Triss - 1 - - 11 2 1 - - - 1 - - 1 - - -------1Cric 1 - 1 - 2 1 1 1 - 1 4 4 - - 1 .................. 1 1 - 1 - - - 1 - 1Poki - - 1 ........................................1 . . i . . l . . -------1 1 - 1 - - -------10Tipu ------------- 1 . i . . | ---------- 1 - 1 - 1 - 1 1 1 ------- -------1 1 1 - 1 - - 1 - - - - - 2 2 - - - -------10Livo ................................................ 1 - - - - - - - 1 - 1 2 - 2 2 1 - 1 - 4 1 -------2 1 - 2 2 1 - 4 110Ouli ...................... 1 - 1 -------110Rise --------------------------------- 1 . . - 1 ---------- 2 1 - 1110Sinni ------------- 1 1 - 1 - 1110Abla . . | ------------- I -------1110Tven ---------------------------] ------------- -------1 - 2 - - 1 ------- 1 1 1 - 1 1 1 -------1110Rhco _ _ 2 -----------------1 ------------- 1 - - 1 -------1111

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 10 0 1 1 1 1 1 1 1 1 1 1 1 1

0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1

0 0 0 0 0 1 1 1 1 1

1 1 0 0 1 1 1 1

1 1 0 0 1 1 1 1

1 1 1 10 0 0 0

1 1 1 1 1 1 1 1 1 11 1 1 I 1 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

1 1 1 1 11 110 0 0 0 0 0 0 1

1 1 1 10 0 1 1

1101

11 1 11 0 1 1 1 11 0 0 1 1 1 1 01

0 0 0 0 0 0 0 0 1

001

0 0 0 0 0 1 1 1

1 1 1 1

229

Tabl

e IV

- 2,

Blan

e W

ater

da

ta.

The

first

maj

or

split

is

iden

tified

by

using

bo

ld tex

t to

sepa

rate

the

tw

o gr

oups

.oo o o o o o o

£ i£ ixcd i« icd i

X iX •

A iCd icd ied i

X ied i

X T“™*Xed 1ed 1

X CNX CO

X iX iX iX iX icd r -H

cd 1

ed —X 1

X 1

X 1

X 1

X CNX CNX coX i

X <NX COX cocd i

X i

X ■

X coX i

X i

cd iXX icd (Ncd CN

X CNX icd i

(N

<N

1 1 1 i i i 1 1 1 1 1 1 1 1

1

1

1

1

1

1

i

i

i i

i i

1 1 -H

1 1

1 1

1 1

1

1

1 1

1 1

1

1

1

1

1

1

1

1

1

1

i

i

i i

i i

1 1

1 1

1 1

1 1

1

1

1 1

1 1

1 1

1

1 1 1 i i i 1 1 1 1 1 1 1 1 ^ H

1 1 1 i i i 1 1 1 1 1 1 1 C N 1

1 1 1 i i i 1 1 1 1 1 1 1 1 -H 1

1 1 1 i i i 1 1 i - H 1 1 1 1 1

— ■ 1 1 i i i 1 ^ H CO 1 1 1 1 CO 1

1 1 1 i i i 1 ^ H CO 1 1 . ^ H CN 1

1 1 1 i i i 1 1 1 1 1 1 1 CO 1

1 1 1 i i i 1 CO CN 1 1 1 1 1

1 1 1 i i i 1 1 CN 1 1 1 ^ H CN 1

1 1 1 i i i 1 1 1 1 1 1 1 1

1 1 1 i i i 1 1 CN 1 1 1 1 1

1 ! 1 i i i 1 1 CO 1 E 1 1 t- h 1

1 1 1 i i i 1 CN CO 1 1 1 1 CN 1

- H 1 1 i i i i i 1 1 1 1 CN 1

1 1 i CN ■ ■ CN CO 1 1 - H . f—H 1

1 I 1 i i i i CN ^ 1 1 a t- h ^ r 1

— 1 1 i i i 1 CO n - 1 1 1 1 <N 1

CN 1 1 i rH i 1 1 i 1 . ^ H ^ 1 - 1

CN 1 1 i 1 1 1 CN CO 1 1 1 ^H 1

i 1 1 i 1 1 1 CN ^H I 1 1 1 CO 1

CN 1 1 i 1 1 i i CO ■ 1 . ^H CO 1

CO 1 1 i 1 1 i i ^ ■ r*H 1 1 CO 1

CN 1 i 1 1 i i ^ 1 1 1 1 CN 1

CN ^ r 1 1 i 1 1 I CO CN 1 1 1 1 CN 1

CN ^ r 1 CN ^H , 1 CO CO I 1 1 1 1

— 1 1 ^ r 1 1 1 CN CO 1 1 1 1 CN 1

i 1 CN t- h CN i ■ ^ ^ 1 1 1 1 x f 1

CO CN t—H i ^ H , i ^ r tT 1 1 1 1 1

CO r - i CO ^ H , i CN CO <—1 1 1 1 1

1 i 1 1 1 1 CO ■ 1 1 1 CN 1

CO 1 i 1 1 I CO ■ 1 1 1 CN

— 1 t- h 1 1 ^ H 1 1 ^ 1 1 1 1 1

CO 1 1 1 1 1 1 1 1 1 1 !—H

1 CN 1 1 1 1 ^ H CO 1 1 1 1 x f 1

1 1 i 1 1 1 1 CN i 1 1 1 1

1 1 i 1 1 1 1 1 1 1 1 1 CNT-H 1 t—h 1 1 1 1 1 i 1 . 1

CO 1 1 1 1 1 1 1 i r -H 1 1 CN 1

CN 1 1 1 1 1 1 "sT » 1 1 1 1-H

CO 1 1 1 1 1 CO i t 1 1 1 t—H

1 1 1 1 1 1 1 1 1 1 1 1 1 1

CO 1 CN 1 1 1 1 1 1 1 1 1 1 1 T—h

— 1 i 1 1 1 1 ^ H 1 1 1 1 1 1 1-H

CN " s f i r -H | 1 1 1 1 1 1 1 1 1

1 1 i i 1 1 1 1 1 1 1 1 t 1 1

> 'S>c xi b ^PQ C/3 f f i P h O PQ W

T3

ocoCN

Esp

a

T“ H

p fi 1

X i 1

■ a 1

A 1

ed 1

ed 1

£ 1

JQ

CM

ed CM

cd -H

X )

cd — *

X> CM

X>

cd 1

Cd CM

X> 1

X> CM

X> (N

X ) CM

X ) 1

X r —H

X ) r - H

cd f - H

cd 1

cd —

X>

X> d -

X ) d -

X> CM

X ) c o

X> c o

X ) CM

X c o

X d "

X

X> d -

cd d "

X c o

X 1

X c o

x c o

X ) CM

cd 1

X> —

X 1

cd d -

cd CO

X d -

X ) Cdcd

Tipu

2

CM

CM

<DCC<

0 O O rH T—H T-H r—i*—i T—H 11 ^ 1 1 t—H

1 1 1 1 fO 1 ^H ^H■ 1 1 CM 1 CM 1 l-H ^H

■d 1 1 1 1 1 1 l“H O

■d 1 1 1 1 1 1 ^H Od rH 1 fO 1 t—H Od 1 1 1 'pH CM 1 O

d 1 1 fO 1 ^H Od CM l 1 1 1 1 ^H O

d CM CM 1 1 1 1 1 O

d 1 1 1 1 CM CM O

d 1 1 1 CM CM ^4 O

d CO 1 I 1 1 O

CM CM ^H 1 I 1 1 O

d~ CM 1 1 1 1 1 O

cn t—h 1 1 I 1 1 O

CM 1 1 1 1 1 1 O r-H

1 1 1 1 1 1 1 O

d 1 1 1 I 1 O■ — 1 I CM 1 O

CM — 1 1 1 1 O1 — 1 1 I 1 1 O t—h

1 CM — 1 I ! 1 O

— CO 1 r—i I 1 1 O * -1 1 1 1 I 1 1 O rH

CM 1 1 I 1 — O

— 1 1 1 I 1 1 O

— 1 1 I 1 1 O1 1 1 1 1 1 1 O

—■ — 1 I t- h ■ O r—h

CM * - 1 1 1 1 1 O r-H

CM 1 1 1 I 1 1 O rH

CO 1 1 1 I 1 ■ O

— 1 1 1 I 1 1 O T“H

CO 1 1 1 1 ■ O

d 1 * 1 1 I 1 1 O

d — - h 1 I sT ■ O-H 1 1 1 I 1 1 O

CM CM 1 1 I 1 1 O t—H

d 1 1 I 1 1 O

d CM I 1 I 1 ■ O

d 1 1 I 1 1 O1 1 1 1 1 O1 1 I 1 I 1 1 O■ 1 1 1 I 1 1 O1 1 1 1 I 1 1 O ^H

1 1 1 1 I 1 ■ O t-h

CO 1 1 1 1 1 ■ O —1 1 1 1 I 1 1 O 0

CM 1 1 1 1 1 1 O 0

1 T—H 1 1 I 1 1 O 0

d CM I 1 1 1 1 O 0

co 1 1 1 I 1 1 O 01 CM 1 1 — 1 ■ O 0

cdu,

X lig I— O

3 <D3M—>D

Whd3> ,

O O s cn O X

COCM

Tabl

e IV

-3,

Dun

eato

n W

ater

da

ta.

The

first

m

ajor

sp

lit

is id

entif

ied

by us

ing

bold

te

xt

to se

para

te

the

two

gro

up

s.

©X ■X iX hH

3 i

OS ■3 i

X i

X ■

3 ■

3 ■

s i ■os ■

x ■03 ■

X r -

X

3 13 1

X CN

-O c o

x iX i

X iX »

X i3 p—H

3 1

3 r H

X 1X I

X 1X 1X CN

X CN

X CO

X 1X CN

X c o

X CO

3 iX i

X iX co

X i

X i03 i

X

X i03 CN

3 CN

X CN

X i3 i

o o o o o © © o © © ©

<N

CN

r*

hJ"

(N - h

m

CN 3 -

cnnj-

_ <D -3 © W) X d <u03 J _ l

CN

c o

O Os ■§ <U XX «

—< CN

1 H

H CO ,

CO ,

1CO

1 1CN ,

1 CN ,

1

1

I iCN ,

1 c o ,

CN CO (

• ^ 1CN CO ,

CN r t (

c o

iCN CO tCN 1

CO %^ ,

^ •CO CN ,

CO CO tCN CO (

,

^ ,

CN CO

CO ^ ,

CO ^ 1,

^ 1

— CO (

CN (

,

f

,

,

,

CN

CN

1

i

1

1

1

1

1

I

I

1

1

1

1 ^

I

. «

1 1 1 1 1 1 1 I l

1 1 1 1 1 1 1 I l

1 i 1 rH T* H rH CO

rH 1 I I r r 1 1 1 rH CN

rH 1 I T f rH H i W

1 I rH I T f CN 1 1 i iN I rH CN <N , i arH I CN I i I i a ^T f 1 <N I T f CO I i a ^CO i — i CO i a aCN i — a CN CN i a aCO I I CN i i i a

i (N a CO -H i i a aCN i i CN 1 a a a t-H i 1 i i 1 i a i i

i <N i , a a a a

— i i rH i — a a ^CN i <N I CN i i aCN i <N I , rH a » a ar“—t i <N I i CN i a a'Tt i i 1 m i a iCN i rH I 1 ■ i a a a

i rH I CN rH a a a i

a n 1 — 1 i a i aCO i 1 CN rH a a a aCO a — a t 1 a a a aCO i — i — rH i 1CN a a CN a a a iCN i r CN 1 a i a a

i CN a CO 1 a a a aCN a CO i 1 i i i a

a CO a CO a a a i

a CN i rH i a a a

i CO i iCN l i 1 a a t aCN rH — i CN <N a i a a

1 i i *—H a i irH i fN i a a i

1 CO a a a a a

1 1 i i 1 a a »—H1CN rH CO i i 1 I i i a

1 CO a i 1 i i a aCN 1 CN i i 1 i i i i

1 CN a 1 a i i a

- 1 © r H

© r H

© r H

© — <

© rH © r H

© rH © rH © -1 © —• © rH © ^

© r H

© r H

© r H

© r H

© r H

© r H

© -1 © rH © ~ H

© rH © r H

© 1 © r H

© r H

© r H

© r H

© r H

© rl© r H

© 1 © © © © © © © © © ©

n © n © n © ©©

i-H i-H

HH ^ O

r-H ^ H O

i—H i—H c ;

H © rH © ©

r H © ©

rH © © r H O O

^H © ©

H © © r H O O

r-H © ©

rH O ©r H © ©

© rH . © r H

© r H -

© - H -

© - H ,

© rH

O r H ,

© r H ,

© rH © r-H

© r H O

© r H O

© © '

© © r H

© © ' © © r H

© © ©

©©

O X > r . 1hco g ~0 S ^ " 0 ‘ 5 - r ^ u r s o 3 P3c/dDCDhO0QPQO

g ^ 3b 3* ■

CUD 3 S <L> fH euoOha XH <CU O•Pi 2 ^

VH o £ *5:-H 3 (U H P»i

232

Appendix V

Appendix V: Individual responses of benthic invertebrates to velocity, depth and substrate.

Introduction

The raw data for the analyses presented in chapter 4 are given below. For each site there is a table giving

each taxon’s log abundance and the corresponding values o f flow variables. Each row represents a

sample. There is a second table for each site giving the results of the gaussian curve fitting.

233

Tabl

e V

-l, R

iver

Et

ive

data

. As

th

e ab

unda

nce

data

is

log

tran

sfor

med

ze

ros

are

pres

ente

d.a

■sc.Q

180 0

OCd

tr o

191 0 ONc-o o

OnO

tnIT,o

oc•c,O o

VC•CNo

•'TCNO

r lO 0

76 cncnO

VC|NClO

Cl«c,O

119 0____

dO O

To

T7o

1C,clO

OOCNo

ClClo

cnCl <NO

clo O

ocOo

rlo 0 0

51 -rro 0 1

2| ClO

ONOO

111 0 ocCNo 0 1

91

£ <NO

,90 0 d 0 1

40 1

6!0.0

00.

29: 810

clO 0.

13 0 06

0.08

100 o

0 05

0.09

' r-o 0 13

0 04 0.12

0 39 r-r-

60 0 0

27

99 0 0.

07 OnClO

NOClO

Clo

890

C I cnClo 0.

59:

0 43

0.34 Clo

dvnCN ■**NO

d O

J;

Subs

trate

Ve

37 voci

9 19 23

125

6

256 (N

VOr-CN

90.7

5'212

.2 r»CO 3723

4.1 On

cic> 123.1

' ■Li •*T£

NO

256,

oe C,Cl 256

98.5 COOCl 73 9

234

1 CN NONCNCl 256 r»

r-TT

.091 NONT)Cl

9 19 25

6 OnCNC- 25

673

973

9'

| Li

mm

us

1 vo

lckm

ari

CN00ci

ood

000 [ f

0 0

0

0 69 o 69

0

OnNOO

690

ood

VOCl

69 0

000 ood

Olig

ocha

eta

1

• rNOci

ooo 1.

79

000 i

ooo

; 0.

69 OOd

OnCl

oo

ooo

’69 0 l.

lOj j

oo

o

[00 0

ooo

T3

00 0

00 0

000000

000 00 0

ood

000 1.10, 000

[o

oo

ood

169 0

ood

000

3a.

Chi

rono

min

ae

1 0

00|

1.79

000 |

000 1

00 0

|

oOd o

oo

j OnVOo 0.00

J

ooo

[o

oo

[00 0

00 0

Ort

hocl

adiin

ae

Ood

I 1 3

9 OnVOd

oo

j 01' l

ood 1.10

1 O ooo

100 0

On Oci 2.

401 1690

Tany

podm

ae

Oil

00 0

| 0.

69

00 0

I

0.69

00 0

00 0

OOO 0 6

91 OnNOd

ood 0.0

01 [o

oo 0 0

0|

1 c

690 i

000 Ooo

O

| 1.6

11

0 00 o OO

d

on oo

dOnVOO

OnVOO 1.6

11

l.lO

j loo 0

OOd

O

Cae

nis

rivu

loru

m

Ood

OnNOO

100 0

ooo

ooo

On Ooo

Bae

tisrh

odan

i!

1 79

I 0

00 OnVOO

00 0

ooo

Oncn 0 691 OnNO

o

169 0

OnNOO

loo 0

oOO

OnNOO

00 0

[69 0

[69 0

[01 l

j P

olyc

entr

opus

690

690 !

ood

69 0

!

ooo

690

ooo

! oo

o

I

OOo

ood 0.

69] [

oo

o 0.69

j

Poly

cent

ropu

sfla

vom

acul

atus

69 0

00 0

ooo

000 8o

ooo

00 0

ood

69 0

_____

0 00

Hyd

roph

la

00 0

|01

I

| 1

95|

2 40

| 1

95

| 1

79 On 69 0

1 \ 1

10 c;Cl

| 2

20f

0 00

| 0

69

00 0

I

On 00 0

| | 2

30

00 0

1

oClcl

234

u€341Q.a.csao<u3e5>41c41UV

G2©-Qs©I*03c/i4>J3>a■<—> c03©£3.2P

©imW)©J.

WDC

ca>£3uCA

30304)

#S41>

wh.©>►

<V © > °. © ° z s * 5

i * 0 o

- f a s ^H Tf- 00 o o OS a s i n cn r~ 1— -H a sCN r- cn ■*t cn s o o OS 00 Os CN r-~ 1—1 00O o o 1 O ^H1 o ' o ' o '■ o ' H1 o ' o ' o ' o 1 o '

so oo CN cn SO SO OO Cn cn CN OO 0 0 o r~-CN r~- i-H H O o so in o CN CN cn SO ino’ o o'• o CN■ o' o' o'1 O■ CN CN■ O o

voo

oo'

0 0 -d" CN SO 00a s -rt- SO m r - in Os _ 00 o s CN o

I—1 cn O CN in T f O CN in in r f s o i-H P—H oo o o'-I o' o' ^H cn’t c n Cn o' o' CN a s■ 0 0 CN c> o' ^H o'1 o'

Id" CN 1—H 00 i n CN SO SO 00 cn N" SO cn cn SO o o0 0 O ■p i n CN O rj- SO O CN cn O cn o oo' o' o'1 o' CN■ cn o' o' o' CNi cn o' o’ o' o' o'

^Ho o o

CN <nc n o

r~o

CNs o

cn••d-

>nSO

r ~o <n

cnCNi n

a s 0 0i n m

mo m

CN o oo^H o o’ o ■'d- SO1 o' o’ o’1 o’ cn K■ o' o’ o'1 o’1 o'

oo

oo

m^H s o

CN<nCN Os

*d-oOs

mo

00CN

cnm 0 0

ocni n r-

0 0OSo

oCN O o

oo

o' f—H o'i o' o'1 o' c n © o’ o’■ in ’ VOi CN o' o' o' o’ o'

CNCN

r-o CN

Osi n

-22.2

7 s oCN cn

c no cn

mcn SO

SOO

Ida s

CNH i nrt"

oo

oo

cn o' o' o'i o’ 1 d o’ o’ o’ o o' o’ o'■ o' c> o’

s o00

00-d-

00o ■Mf

a s r~ r -©

CNCN

cnCN

Os00

CNO

OsOO SO

cn r -CNCN

SO cno

oo

o' o' o'1 o'1 o' o' CN o’ o’ o’■ o’ cn■ o' o' o'■ ^H

1 o’ o

t"00

a si-H

a so

i n oSO CN

©cncn

ooo

oH SO

r ~^H 0 0

p000 0

cn -d- i nSO

oo

oo

o' o' o' o’i CN O o' o’1 o’ i 1-H o' o' O■ o’ o' o

("o

cna s

r~ 00CN

00o Os

cn po■d- o

oso’d"

i-Hcn

i n"d- "d-

i nCN^H

Os1-H

r~-cN

^Ho

oo

o' o o■ o’ o’■ o' ^H © o’ o’ o1 O■ CN o' o' O■ o’ o'

CN00

CNCN

Ht1-H

SOCN

OSs q

oF“H

00r - o

oooo

cn00

i nr- Os

pi n^H 00

^HCN o

oo

oo

^H o' o' o’■ cn i n■ o’CN © o’ o’ cni c n o' o' o'■ o’ o’ o'

omCNso

SOO

cnC"

Os00

op SOo

rcn

^Ho >n

(N\D

soO cn o

r -cn o o

oo’ o' o'1 001 i n

CN o’ o’ ^H cni ^H o' o' o’1 o’

cd 

*o OCm d O -—i CN X> X>

235

Tabl

e V

-3,

Dun

eato

n W

ater

da

ta.

As

the

abun

danc

e da

ta

is log

tr

ansf

orm

ed

zero

s ar

e pr

esen

ted.

309

>

(Ua35ooe>o

«oT3

3CX

<o3T3

eO3O«-i2u

•I |I so ■§

ss:o

Q

<*>-30&1£'o ir*a qq"S' |c c

oo

oo

i n in i n. CD CO 00 r- .

CO CD CDN" N" m N -

CD N" CD COCN CN CO x—O O o o o

CXD CD h - CO CNCD lO CD CN X—

O O X“ o oO o o o o

CD o CD oCO o CD oO o o o

s cn co inL Ji O 0> 0> 2CN CN N" CD

CDN-COCNLOCNCNLOCOCNN;N-CNCOCNN-d o d d d d d d

CD S O) N O) O) CO T - CO T - T -

CD lO N" x-o o

O O O ^ C O T -C N O M in cO( Ot- cOCDCOCOCOt-CNM-v; O t- t- O O O t-O C Nd d d d d d d d d o

h-oo

h-CNd

cocoCNo

in t- t- o co CN o d d

o CNo d

00 CD in 0 0 ino CO 00 d 00CO h - CD CO CD

CD r - CD CD COo X— X— o COo d o o d

CO T_ o CO CDCD CO 00 CD CNX— CO CO T — CNd o o o o

o CDO CDd o

ooo

O O CD CD O O x - c o CO o

§ °> LO CO

O CN COo o o

O CD x- (O C\l t CN CO CNO O O

oo

o o o o O o o 00 CD O x— O O O X— O O Oo o o x- ; O o x - N" CD O CD x— X— O N- N" O Oo o o x - d o x~ CN O O X— X - X— o CN CN O o

X— N" o CD o CD N - O O h - CN O in O CD CD oCD N - o CD CO CO CD NT O CO CN O CO CD O CD oCO co d o CO X - CN CN o c o CO CO CO CO o o Csi X - 1

o o X— X - CD CD CD OX— o CD CD CD CO CO ■x—X— o X— x - d d X— x—

CDCDo

oo

CD O CD O CD X—CD O CD O CD CDd d o o o X—

CDCOCN

o CD X— X— o o CO O o CDo CO CD CD X— o 00 o o COCO x— X— X— XT" o CN CO d X—-

CD N" h - o ON - x— ■x— N" CN

T“ CO CN CN

CD CD o o CD CD i n o o ^ r CO 00 CD o CO 00 00 o oc o Is*- CD o o CO CD a> o q o CD CO o CO o 0 0 o o N"

T_ x - o o d T“ O T“ o T_ o CN CO CN T~ o csi CN CN cvi T_

OO OO CD COo o

oo

o o o oo d

CD CD o oCD CD ■X— o < Dd d X— o

oo

CDCO CDCDCO o

oo

CD CD CDCD CO r-*d T_ T_

o o CDCN o COCN o d CN CN O O

ca(U3

>

03OobO

<D0323ex

5:I ij►o O

I ^ 05 CO CO Is- 05 ini f x - CM r CM in 05 __ CO in CM M" (O _ x_ CM CO CMSi ° x— CM o CM X— No CM CM o x— o . X~ CM o x— X~ o

i ° d o o o o O o o d d o o ° o o o o o ° o d o o

CO CM in M- 05 05 00 05 05 CO 0 5 CO Is- 00 M" CM Is- T_ o Is- CM 05 COCM X— CO CO CM 00 CO CO X— X- Is- CM M" o M" CO 00 X— Is- x — CO m COx- CM 00 O CO o X— o o CO CO X— o i n CO CO X— i n CO CO X— o

d O d o o o d o o d o o o o o o o o o o o o o o

O 05 05 05 o o o o o 05 o o oX- CO CO CO o o o o o CO o o oT_ o T— o d d d d o d o o o

o o o in 05 X— 05 05 o o 05 o 05 o 05 05 o o 05 COo x- o 5 05 CO CO CO CO o o CO X— CO CM Is - CO X— CM CO ino o o o d o o 'r ~ CM T— o T_ CM T_ CM

CX) in 05 o o 05 05 o 05 o o 05 0 5 05 o 0505 CO o o Is- CO CO s CM CO CM o CO CO CO CO CO

CM o d o' CO X- d CM CM d d d CM d

*- s5 s: bo

-s: ~

■3-

«ss

I

s* >

CM 00 M;K. CO CO fv CO in CO in in m

c dCM

CMIs-

inCO

1 ^CO

CO05

1 —CO X—

CO0005 CO

0005

0005

0005

05 O LO CO 0 0 00 * - 05 CM

Is-h-J-

00 o un° o Zo ^ ôo CO 0 0

T_ 0 5

0500CM

LO05

o 05 o x— in o o 05 o 05 05 X—o CO X— CD 05 o o Is- X— CO CD CDd X— X— X— X— d o X— X— d d X—

m in o CO CO x— o o X— 05 05 m in 05 x— X — 05 05 o05 05 M" 00 in CD CO CM CO o CO CO 05 05 Is- CO CO CO Is- COX— X— CM CM CM X— csi c \i X— cd o d X— X— X— X - X— d X " CM

05 05 co o 00 o o o X— o 05 o 00 o 05 o 05 X— oCOCOo x— o o x— x— Is- ■sf CO ■M" CM COCOX— COIs- X—o o CM X— CM o x-^ X— cd CM o CM csi CM d CM X— o CM x -’

05 o 05 O 05 05 05 o 05 05 05 05 05 oCO o CO T - CO CO CO CM CO CO Is- CO Is- CO oX— o X— x—" o o X— CM CM d X— o X— X— d

OO

05COo

o o o od o

oo o o o O T- o oo

oo oo

X— 05 o o CO o o o 0 5 05 o 1 ^ o o CO CM 05CO Is- o o o o o o CO CO CD CM Is - o CO M" CM COX— X - o d CM o o d d X— CM CM CM d CM CM cd o

237

cC3

3Mc7i

aftW)aftCUDc

cCit3uC<3

3«COC

«■oft-fta

*c0-ta-ceas3

Q

h ?1>Ci

3«

H

cd-*-*<Dcdoo0 0

CDcd

TD

3CX

(Dcd

TD

SoGOft

2u

ao

cna

•SftoX.ca

< 5ft

s sJc?s -S•*** *"1*O o

X

«nftv.ftft0

1 -ft:

I .u"§h£ r

ca«Qq<a

- f t0£>

§ •

1cn ft5ft ftq>-«4 -Ki

ft

Io f t <a

<1

*§J

cn m £7 'O M ^ NO m ° .O O 9

- s(NO

''D-on

CN so00 o o^r o cn00 o o ’

CN3 -i n ©

©

m ON t—h

3- CN ONCN <—< oCN o ’ o ’

ON ^r cnCN o f"NO m i—11—i o ’ o ’

csi a S!^ 3 ^ ° d o

oo r - Os so IT)

j ^ oo o

^ 2 ° l A l V r — 1

o’ Q o’

m 00H*T <cn 00NJ

CNo ’

CNoo ’

r -o

CNo ’

cn! i

cnSOCN

t—h COA CN,T ocn o o ’

V i W im tN oo’ cn

r - CN ON 00 CN3" o m On r-Ho ’ o ’ o ’ o ’ o ’

inO in om 00 m

o 'To ’

003 ’ cn 3 d (N 7

sot"-

O m00 cn r->o ’ f—I o ’

u . z i 2 td0J ■•-> £ “

■Si £ -a s 5 « 2.. _ 3 g_ _ <*> - ft,c , H O ’—1 CNX pH K cd JD X) X)

in m .ON T-l ^Tf- so o o’ o’ '

3 - NO cno ON r -m o oCN o ’ o ’

3" NO in NO OnON H On cn00 CN l"- r —H o

1 T“ H t—h o ’ o ’

CN so r -

cn o

h m 2so cn p: O On ° .o ’ o ’ 9

r -H C N

ON NOm m o' o'

o o OO r -NO On r - CNo ’ o ' o On

i i o ’ o ’

3 -

3 -

cnin r -

NOr -o CN

NO00

CNm* m NO CN i-Hoi 00 ■ o ’ o ’

___ On inCN [-- NOo ’ CN NO1

vo ^ 00 t t ^ ^, ’ CN r~ l

r--o

3-cn

mCN

ON NO rÎ"1 c i—io ’ 1—1 CN

cn r - cnCN C"

i—iT|-NO00

cnin 00 r -

o oo cn o NO 3"r -H 00 r -H• o o ' o ’■ CN

^ r- 9 m cn 00

cn m NOo m 1—1o ’ o ’ CN i

^ a “ •f" cn

r-~CNo ’

cn rr-mON

NO NO m O n r o

m ON ONcn o ONTT i i oCN o ’ o ’

r - cn1-H in inm ON oo o ’ o ’o ’ 1

oo 00o OO oor - o oCN O O

no m cn r— 9 ooo ’ o ’ 9

NO ON 1— 1 ONo

NO r - cn m cn00 NO in cn "d" ON o cnm m CD CN cn Oo ’ o ’ o ’i

t- hi

oi H o ’ o ’

m oo nom r~- ^v > c ”cn r - o ' o '

3-o

cn r - r -O cn o

3- i—icn o o

CN o cnCN NO TfO t—H CNin o ’ o ’

cn°N £ 2 NO .O ’ ° 9

r- __

g gcn 1 . O’ °

inCNo ’

NO m 00 cnr -

00 mON

t|- cnTf3"

tTNO

cnO ON o NO

oCNH

o ’ o ’ o ’■cni m cn o ’ o ’

aCiO

o"ScdooPh

uy£ . 2 3■*“ 4—> 4—ICDJ33>cu

T3<D _ ,ft 3cd o—t -t-<3. c r pon -G,TO O i—i CN

K cd X> Xi X>

Ci-*■*esftcn,£3(/)

O333PU|

CD_3cd>CU

TD<D __, ft TD 3 <D

. TD O «—■ CN « cd JO JO X>

0.92

-2.02

-8

.44

-0.05

0.8

3 1.9

2 -0

.03

1.33

2.77

1.00

0.78

0.00

0.08

0.14

0.02

0.01

0.00

0.03

-0.02

-0

.03

0.00

-0.0

10.0

0 0.0

0 0.0

0 0.0

0 0.0

0 0.0

0 0.0

0 0.0

0 0.0

0 0.0

0 0.

00

Tabl

e V

-5,

Blan

e W

ater

da

ta.

As

the

abun

danc

e da

ta

is log

tr

anst

orm

ed

zero

s ar

e pr

esen

ted.

l-l -t—< ON c n o c n un 00 i n NO N" cn o N" o oo ^ r c n t" - CN NOJO c n ^ r c n 'vf cn c n c n m c n CN r~~ 1—I CN CN 1—1 NO NO in 00Don

JS <N ^ r ON o ON CN O n ON ON c n NO CN c n r-~- ^ r r -H NO ^ r o cn 00e xD B CN o 1— cn CN o o O CN '1 CN CN o 1 CN i—1 CN o oQ o' o' o’ o' o' o' o' o' o' o' o' o' o’ o’ o' o’ o' o’ o’ o’ o' o'

■4—*r - IT) c n ON o o 00 ON c n cn c n o 00 m c n OO i n 00 m c n o o O n o io ,c i—i c n NO o o C-" N" o ON c n oo T < ■*d- CN CN c n CN CN o o T—H i no C ^ r i n o o o c n m NO 1-H i—1 o c n c n c n CN l—1 m c n CN c n o CN m"a>

>G o 1 o' o’ o' o’ o' o’ o' o’ o' o’ o’ o' o’ o' o’ o’ o’ o’ o’ o'

•S Q o o ON o o O o o ON ON ON ma r* o o NO o o o r — I o NO NO c n O n

3 5J o' o' o' o’ o' o’ o’ o’ o’ i—H

a fc .a ©S J i

©

aOtoa'W.Ci,bo

©S 's a ^<5 o Cn .aS O3 03

a*^ afr-g a &

§oa<1

■a s a I© o "Q-S

toa!kaa©

-aa.

o oO T-H

ooo'

oo

onr -

VO O Oo o ONNO ooo o o

oTfCN

NOONNO

oo O nr-~

o o o oo o ’

oo

oo

oo oo

oo ON O n c n c n

ONc n

cncn

O n m NO ON

oo

o o

ONc n

ONc n

oo ooCN

00o<N

O o o o O n O n ON ON o o O n 00O >— _ CN o NO c n NO NO c n i— • o c n oo ’ CN o ' o ' r ~ l o ’ < I o ' r* ^ CN

O nNO

oo

cn

oo

oo

ON o o in r -NO o o O n r-o’ o’ o' T 'H CN

O n 00 ON O n O n o in ON o m O n O O n ■*d- i—ir - ^ r c n r-_ c n <—; ON NO o ON r-~ «“ ; r - * NO

i—I CN i—’ r —H -—I i—i o' o’ r —H l—H c n i—i

0 0 O n r - 00 t- - O ^ r c n ON ON 00 o o 00 mN" c n c n m oo On o 1—1 c n OO OO o r - m CN ON

"cT c n c n CN c n c n c n CN CN CN CN CN c n ^d" c n

CDCOCN

-O3tn

o- c 

r i ^

S ?5

•« ^ ^ O >

0t-n ■>*

&C

Q1 *I ■§8 s

■2 R1 1 0, .g

>)aaco

f3

-5:

m CN cn NO OO CN in ,—1 ON r-" r - t"- r - oo in On r -n r 3" N" 3" CN NO CN Xj- cn cn cn cn CN CN n - cn

on o cn NO r - N" in 3" NO NO O NO tJ-,—1 !■■ 4 f-H o 1—4 >—• cn o i—1 o O o o 1— O i—i oo ' o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’

,_, in o cn CN ON in ^r cn cn cn 3- ON CN cn r fm> NO cn On 00 3- CN NO oo cn i—i 00 oo NO m m inr - o ON cn oo ON 1—4 o NO cn cn cn m NO CN NOo ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’ o ’

On o On ON o o m ON ocn I—1 NO NO o o NO On NO cn

1— o ’ o ’ o ’ o ’ T— o ’ CN

O 3" O o ON O o oo o ON o o ON<N ON O o NO o o ''3- o NO CN o cnr s CN o ’ o ’ o ’ o ’ o ’ CN o ’ o ’ CN o ’

o o On ON o ON ON O o o’—_ o NO cn o NO NO NO CN o o NO

o ’ o ’ '—, o ’ o ’ 1—1 o ’ CN o ' o ’ 1—1

o ON o o'—« cn o o

1—1 o ’ o ’

1" On 00 m ON o On ON ONno cn o ON r - CN NO NO NO<Ni

"CN

”CN o ’ o ’ o ’

On On o o O O oVO NO o o cn »—1 oO o ' o ’ o ’ CN 1—1 o ’

ON ON ON f 4 NO N" On ON ON ON ONNO NO cn t"- m NO NO cn NO 00 r - cno ’ o ’ 1—4 CN CN CN T**H 1—1 CN ' '

o m o oNO NO o ON o o

o ’ o ' o ’

o NO o o oo o o ON1—' m NO r-H CN o o o cn

CN 1—I CN CN o ’ o ’

O cn ON 3- ON NO o cn ■ r ON On 3 -m cn 00 m r - CN m oo NO NO r - r -^r 1—H cn I—i cn «—<’ cn cn cn cn o ’ 1—4 cn

240

cOJD

i Z

aLa01)au

0£c

1=<D

t3uCa

3C8Oi

«S■oLa4>+HCS

£a>3CO

5VOI>

2etH

cn QI =fcq §

•N*o O

aO

cnSS>•*»

t>C

P Ocn8 O5 05

U5 •*«-I ^ ^ ■ S* sJ S

INjacu

< l

cn R

* -S, 5 - 5Cl

O N CN NO i n m r— c r ~ ' — • O m O N oo 3"3" N O o 3 " o CO O N 3 " 9 O N 3 " o r- 3- o oO o ’ o ’

1 o' o' o’ o' o' 9 o' CN■ o' o' o o 9

r -m

C OC N

3 ÔO N3 "

OO C OC N

9 o’ o’ o' o' 9

oo ON NO IT) o O nCO NO o oo o To' o o’1 c> o’ o’

oo cn r? no r r\o ^ a °O O 9 9 9

£ •* 3 .2£ ta> pH

onNO

<l>3

1 3>

PH

T d«-H33crcn

P h

NOO

r -vo^ ^ 00ov 9 o\

o ' 9 o ' <9oo

no 3* 9 o o oO On ° . Os r. OO O 9 o' 9 o'

NO CN ON NO CO o r- Onin r"H r—> o r“ H CO © 3"CN o' o' o'■ o' o’ 9 © o'

C Oc o c o

oo

CNON 3"

oo ° 9 ooo

00 o in CN NO 00 NO o o O 00 NO c n CO00 3- 3 " CO o CN CN r " o o CO o CN 1 NOo’ o’ o'■ o’ o' o’ o’ o' r—H 3 -' o’ o' o' o'1

o3 "

NO NO 3 " o NO m 3 " OO 3 " oo o C O oo3 ; C N O o o o NO CN o c o r - o oo’—l o ' o ’ 1—11 o ' o ’ o ’ o

1 T—' o ' o ’ o ' o ’

3 - N O C O ■'T 0 0 i n 3 - 3 " m o C N ON3 ; C N ■3" O C N r - C N o oo o T- ON

C N o ' o r—41 o ’ 9 9 o ’ o o ’ 9 o ' o ’ o ’

r-o

ox 1—1 CN X X

CN3"

3

'O<D

333

33a>Wh33cr

P h P h P4

r-o

oino o

l > r - £ r O n c nCN Is" 9 0 0 3 "o o’ 9 o' o'

in o in in CN m t - OO CN ON o ON in CO NO o00 CO CN in o in CO 3" "“I NO o r- m *“! 3"9 o ' o ’ o '1 o ' 9 CO o ' o o1 9 o' co o' o' o■ 9

CNCOo'

OO O ) Ono 9 3 " m

o' o 9 o o'

NO

oo’

ooo

N OC O o 3 "

0 0

3 -O N

Oo O N

C N3 - O m o

3"T—H o ’ o 9 C N■ o ' o' O o’

■ 9 o'

C NO

ooC O

ooin3;

C No OC N

On (NC O

r —1o NO

C Noo

r-NO

C Nm

mo

0 0C N

0 0C O

9 o' o' < N o' o’ 9 o' o' C O C N1 o' o’ o’ o' C O o’

oX CN

X

-7.30

-8

.66

-2.19

2.0

6 7.5

3 -2

.36

16.83

-2

.09

-5.30

-1

0.50

Documents

Flow preferences of benthic macroinvertebrates in three